1Report T39-T4 August 2013 The Potential and Challenges of ...task39.org/files/2014/01/Task-39-Drop-in-Biofuels-Report-FINAL-2-Oc… · (collectively known as hydroprocessing) is

IEA Bioenergy

1Report T39-T4 August 2013 The Potential and Challenges of

Drop-in Biofuels

A Report by IEA Bioenergy Task 39

AUTHORS:

Sergios Karatzos University of British Columbia, Canada

James D. McMillan

National Renewable Energy Laboratory, USA

Jack N. Saddler University of British Columbia, Canada

Report T39-T1 July 2014

http://www.iea.org/topics/cleanenergytechnologies/ieaenergytechnologynetwork/

“The potential and challenges of drop-in biofuels” IEA Bioenergy Task 39 ISBN: 978-1-910154-07-6 (electronic version)

July 2014 1

The Potential and Challenges of Drop-in Biofuels

Report prepared by: Sergios Karatzos University of British Columbia, Canada James D. McMillan National Renewable Energy Laboratory, USA

Jack N. Saddler University of British Columbia, Canada Copyright © 2014 IEA Bioenergy. All rights reserved. First electronic edition produced in 2014 A catalogue record for this Report is available from the British Library. ISBN: 978-1-910154-07-6 (electronic version) Published by IEA Bioenergy

Disclaimer

IEA Bioenergy, also known as the Implementing Agreement for a Programme of Research, Development and Demonstration on Bioenergy, functions within a Framework created by the International Energy Agency (IEA). Views, findings and publications of IEA Bioenergy do not necessarily represent the views or policies of the IEA Secretariat or of its individual Member countries.



July 2014 2

THE POTENTIAL AND CHALLENGES OF ‘DROP-IN’ BIOFUELS

EXECUTIVE SUMMARY

This report was commissioned by IEA Bioenergy Task 39 with the goal of providing a background to the

topic, an assessment of technical approaches being developed and an overview of anticipated

challenges in large scale commercialization of so called “drop-in” biofuels. For the purposes of this

report, “drop-in” biofuels are defined as “liquid bio-hydrocarbons that are functionally equivalent to

petroleum fuels and are fully compatible with existing petroleum infrastructure”.

The global petroleum industry is expected to require increasing amounts of hydrogen in the coming

years to upgrade crude oil feedstocks of declining quality (i.e., increasingly heavier and more sour),

particularly in areas where especially heavy oils are being sourced such as Venezuela and Alberta. For

the foreseeable future, much of this hydrogen is likely to be derived from natural gas. At the same time,

there will also be increasing demand for hydrogen to deoxygenate biomass (carbohydrates and lignin) to

produce drop-in hydrocarbon biofuels.

Oil refineries use hydrogen to upgrade low grade crude oil by removing sulfur and other heteroatom

impurities (hydrotreating) and by “cracking” longer oil carbon chains to shorter chains while also

enriching them with hydrogen (hydrocracking). One result of these hydrogen-consuming processes

(collectively known as hydroprocessing) is to elevate the hydrogen to carbon ratio of low grade crude

oils. The hydrogen to carbon ratio in petroleum feedstocks is a good indicator of their quality for fuel

production since a high sulfur content as well as the presence of long and condensed carbon chains

(e.g., in coal) reduce the H/C ratio. As detailed in the main body of the report, the H/C ratio can be

visualized as a staircase in which the more “steps” that have to be climbed up the “H/C staircase”, the

more hydrogen inputs and processing efforts are required to elevate the H/C ratio to the level required

for higher grade liquid gasoline, diesel and jet transportation fuels. Non-hydrogen-consuming processes

such as catalytic or thermal cracking can also improve the H/C ratio of petroleum feeds by removing

carbon in the form of tars and char (coke). However, this approach consumes feedstock and reduces

yields and so is generally avoided, particularly when crude oil prices are high.

It is also evident that a majority of evolving drop-in biofuel technologies require hydrogen (H2) inputs or

other chemical reduction processes to upgrade oxygen-rich carbohydrate, lignin or lipid feedstocks to


July 2014 3

hydrogen-rich hydrocarbons that are functionally equivalent to petroleum-derived liquid fuels. As

detailed in the report, a variation of the hydroprocessing step will likely be common to many drop-in

biofuel technology platforms, with imported hydrogen used to remove oxygen (in the form of H2O) from

oxygenated lignocellulose intermediates or lipid feedstocks. Alternatively, non-hydrogen consuming

processes (whether chemical or biological) will have to oxidize significant amounts of feedstock carbon

in order to produce the required hydrogen or alternative reducing power carriers (e.g., nicotinamide

adenine dinucleotide phosphate or NADPH). However, these alternative routes to deoxygenation are

generally less attractive as they can consume a significant amount of the feedstock. After adjusting for

the oxygen content of the biomass feedstock, the hydrogen to carbon ratio, Heff/C, can be defined as a

relevant metric for drop-in biofuel processes. Highly oxygenated biomass feedstocks such as sugar

molecules have a Heff/C ratio of 0 whereas the target for drop-in biofuels is approximately 2, similar to

the H/C ratio of diesel. Most biomass feedstocks (sugars, biomass, lignin) have a low Heff/C ratio and are

thus situated near the bottom steps of the H/C staircase. Biomass feedstocks thus need to “climb” more

steps than fossil feedstocks to reach the chemically reduced state of diesel, jet and gasoline fuels. Even

low grade fossil feedstocks such as coal (H/C = 0.5) contain a substantially higher Heff/C ratio than most

biomass feedstocks. A notable exception are the biomass lipid fractions and other renewable

oleochemical types of feedstocks, which contain much lower levels of oxygen and have an Heff/C ratio of

about 1.8 and are thus much farther up the H/C staircase and more readily suited for conversion to

drop-in biofuels.

There are several ways to produce drop-in biofuels that are oxygen-free and functionally equivalent to

petroleum transportation fuels. These are discussed within three major sections of the report and

include: oleochemical processes, such as the hydroprocessing of lipid feedstock from either oil crops,

algae or tallow; thermochemical processes, such as the thermal conversion of biomass to fluid

intermediates (gas or oil) which are then catalytically upgraded/hydroprocessed to hydrocarbon fuels;

and biochemical processes, such as the biological conversion of biomass (sugars or cellulosic materials)

to longer chain alcohols and hydrocarbons. A fourth category is also briefly described that includes

“hybrid” thermochemical/biochemical technologies such as fermentation of synthesis gas and catalytic

reforming of sugars/carbohydrates.

To date, oleochemical based processes have been the main supplier of the drop-in biofuels that have

been evaluated for commercial application by sectors such as aviation. These processes require a simple


July 2014 4

hydroprocessing step to catalytically remove oxygen from the fatty acid chains present in the lipid

feedstock to convert them to diesel-like hydrocarbon mixtures. This technology is well developed, is

maturing and entails relatively low technological risk and low capital expenditure compared to other

emerging drop-in biofuel production routes. Most lipid feedstocks have relatively low oxygen content

(11% mass) and thus require lower hydrogen inputs to be upgraded to liquid transportation fuels.

However, the feedstock is generally costly and available in limited supply, as vegetable oils such as palm

and rapeseed are currently priced in the range of USD $500-$1200/t (or $12-30/GJ) compared to

approximately USD $75-$125/t (oven dry basis, or $3.75-6.25/GJ) for lignocellulosic biomass, and their

supply is often limited by competition from other value-added end users (e.g., food and cosmetics

industries). There are also ongoing challenges regarding the sustainable production of vegetable oils as

production is relatively land use and resource intensive. Although “food-vs.-fuels” concerns and related

debate are likely to continue, several companies are operating commercial oleochemical feedstock-to-

biofuels facilities around the world, including Neste Oil (Finland, Rotterdam, Singapore) and Dynamic

Fuels (Louisiana, USA).

The various thermochemical methods currently being assessed for biofuel production have their origins

in the ancient process of “burning” biomass in the absence of oxygen to make charcoal, a higher calorific

value product. Thermochemical processing conditions can be optimized to influence the ratio of the

three main products of bio-oil, synthesis gas and char. The two main routes to drop-in biofuels are

through pyrolysis and gasification. Fast pyrolysis (essentially treating biomass at 500 °C for a few

seconds) has been studied in detail since the early 1980s and bio-oil yields of up to 75 wt% can typically

be obtained from various biomass feedstocks. Although there are a few, niche high value markets for

bio-oil components, such as food flavouring (Barbeque flavour), today pyrolysis liquids are primarily

considered for use in stationary power generating facilities such as the proposed 720 tpd Pyrogrot

facility in Sweden. Bio-oils can also be upgraded to drop-in biofuels although this requires significant

hydrogen inputs. While these hydrogen inputs can be generated from the biomass feedstock itself, this

process is inefficient when compared to sourcing hydrogen from an external source such as natural gas.

The pyrolysis platform requires large amounts of hydrogen gas inputs which represent a large

proportion of both capex and opex in a stand-alone facility. It has been estimated that sourcing external

hydrogen from an oil refinery can reduce the capex of a pyrolysis drop-in biofuel facility by ca. 40% and

the opex by ca. 15% (Jones et al. 2009). Pyrolysis platforms also have great potential to leverage oil

refineries in order to reduce biofuels production capital and operating costs. The major saving will in


July 2014 5

part be a result of sourcing hydrogen from the oil refinery directly. However, it is estimated that current

US refinery hydrogen capacity of 3 billion standard cubic feet per day, would need to be tripled to meet

the 2022 US RFS cellulosic advanced biofuel mandate of 15 billion gallons (57 billion L) using pyrolysis

platform-derived diesel/gasoline blendstock. Although existing hydrocracking units (downstream in a

refinery) can co-process petroleum and hydrotreated pyrolysis oils (HPO), this practice is not yet

commercial and it comes with challenges related to adapting the catalyst design to accommodate two

disparate feedstocks (HPO and petroleum). A case study where Haldor Topsoe (the world’s biggest

manufacturer of petroleum refinery catalysts) performed trials on industrial hydrocrackers using various

biofeeds identified several challenges to catalytic “co-processing” of biofeed blends with petroleum.

Although further upstream insertion points have been suggested, such as at the vacuum distillation

tower, these alternative processing strategies can only be used with minimally processed pyrolysis oils

which can contain large amounts of refinery contaminants such as oxygen and inorganics. Two of the

major challenges constraining development of pyrolysis derived drop-in biofuels are the availability of

low cost sustainable hydrogen and the technological advances needed to adapt hydrotreating catalysts

to bio-oil feedstocks. Various companies such as Canada’s ENSYN have operated pilot plants for several

years and KiOR recently completed a 49 million litre per year (MLPY) or 13 million gallon per year

(MGPY) commercial facility in the US.

The other major thermochemical route to drop-in biofuels is through gasification. Gasification of

biomass or bio-oil produces synthesis gas (“syngas”, comprised of mostly H2 and CO), which is primarily

used to fuel stationary heat and power facilities such as the 8 MW bio-power station in Gussing, Austria.

Syngas can also be upgraded (catalytically condensed) to drop-in liquid biofuels via the Fischer-Tropsch

process (FT), which has its origins in the 1920s in Germany when access to oil was problematic. Since

the 1980s, South Africa’s Sasol converts coal syngas into diesel at the CtL Secunda facility which has a

capacity of 160,000 barrels of diesel per day. A variation of the FT process is used in the world’s largest

natural gas-to-liquids facility (Shell’s Pearl GtL facility in Qatar, completed in 2011) to produce 140,000

barrels of diesel per day. However, biomass derived syngas is less energy dense than natural gas and it

contains more impurities and a lower H/C ratio. As a result, biomass syngas needs to be enriched in

hydrogen and cleaned of the impurities such as tars, nitrogen and other heteroatoms that can

deactivate synthesis catalysts. Hydrogen is typically produced from the syngas itself by a process known

as the “water-gas shift” reaction. However, this reaction consumes feedstock carbon and thus reduces

the overall biomass-to-fuel yields. Alternatively, as is being proposed by companies such as Sundrop


July 2014 6

Biofuels in the US, the hydrogen can be derived from natural gas. Generally, gasification technologies

entail high capital costs to both gasify the biomass and convert the resulting syngas to Fischer-Tropsch

liquids or partially oxygenated liquid hydrocarbon products such as mixed alcohols. To benefit from

economies of scale, these types of facilities usually have to be constructed at large scales. The capital

cost estimates for a first-of-kind gasification-based facility are in the region of USD $600-900 million.

Several companies are pursuing gasification platform routes to drop-in biofuels such as Forest BtL Oy in

Finland, which has licensed Choren’s Carbo-V technology and intends to complete a 129 million litre per

year (MLPY) {or 34 million gallon per year MGPY} facility by 2016.

The capital costs of both the oleochemical and thermochemical processes could be reduced by

leveraging existing process units available in petroleum refineries. Oil refineries are complex facilities

comprised of the many unit operations needed to fractionate and upgrade diverse crude oil feedstocks.

Upgrading entails a number of intertwined processes such as cracking (breaking heavy hydrocarbon

chains to lighter ones), naptha reforming (creating aromatic molecules necessary for gasoline blends)

and hydrotreatment (mainly used to remove sulfur before fuel blendstock finishing). The dilemma in

trying to identify refinery insertion points for renewable feedstock drop-in biofuel intermediates is to

what extent should the intermediate be upgraded (deoxygenated) prior to insertion and to what extent

should the refinery be adapted to accept less-upgraded, oxygen-containing biofeed intermediates. The

challenges of processing biofeeds in an oil refinery are significant, as has been demonstrated by

previous industrial trials using less problematic renewable feedstocks such as fatty acids containing

relatively low amounts of oxygen (11% oxygen). The oxygen content of biofeeds translates to corrosion

of metallurgy and extensive coking of catalyst surfaces as well as downstream contamination risks and

requirements for venting of oxygenated gases (CO, CO2 and H2O). Strategies to mitigate these challenges

include limiting the blending rate of biofeeds in petroleum feeds and favouring insertion points towards

the end of refinery processing, both of which lower the risk of downstream contamination with biomass

oxygenates, inorganics and tars. Hydroprocessing units situated at the end of the oil refining process are

suitable for drop-in biofuel leveraging. All of the drop-in biofuel processes proposed to date entail some

form or degree of capital intensive and hydrogen-consuming hydroprocessing (especially pyrolysis and

hydrotreated ester and fatty acids (HEFA) platforms). Refineries can be leveraged by drop-in biofuel

facilities in order to utilize existing hydroprocessing facilities and also to source low cost fossil feedstock

derived hydrogen. Still, even with this lower risk co-location strategy, there are significant challenges


July 2014 7

that need to be resolved such as matching the scale, siting and catalyst design for two distinctly different

feedstocks (bulky and reactive solid biomass versus relatively inert petroleum liquids (crude oil)).

Biological routes form the third category of drop-in biofuel technologies. These include metabolic

pathways that convert highly oxygenated, low Heff/C, sugars to high energy density molecules such as

butanol (e.g. Gevo, Butamax), farnesene (e.g. Amyris) and fatty acids (e.g. LS9). The metabolic processes

involved in biologically deoxygenating carbohydrates to drop-in fuel molecules are energy-intensive and

they are usually employed by the microorganisms when under stress and as mechanisms to store energy

or build defence barriers (e.g. lipid layers). In industrial practice, this generally translates to biological

systems with low volumetric productivities and less stable metabolic pathways. These so-called

advanced fermentation pathways are not as efficient as conventional sugar-to-ethanol industrial

fermentation systems. A key advantage of biological compared to thermochemical routes, is their ability

to produce relatively pure molecular streams with predictable chemistry that can be readily

functionalized (chemically). Thus this route can take advantage of the rapidly growing value-added

chemicals and polymers markets. These markets consist mostly of organic diacids and dialcohols

(butanediol, succinic acid etc.) which have lower Heff/C ratios than hydrocarbon-like drop-in biofuels.

Thus they are "easier" to produce with fewer processing efforts and fewer hydrogen inputs. In the near

term, it is likely that the biological platform will exploit the higher margins that can be achieved in value-

added biochemical markets rather than fuel markets. Various business intelligence organisations have

estimated significant growth for these bio-based chemicals over the coming decade (e.g. ca. 20%/year to

reach 50 million metric tonnes by 2020). Until these lucrative chemical markets are saturated, there will

be little incentive for biological conversion companies to produce biofuels.

A fourth category of “hybrid platforms” combines elements of the categories described earlier. These

include fermentation of syngas (example, LanzaTech), alcohol-to-jet (example, BYOGY), acid-to-ethanol

(example, Zeachem), and aqueous phase reforming (example, Virent). Each of these technologies has

certain advantages such as improved utilization of feedstock carbon (Zeachem, LanzaTech) or use of

commodity bio-feedstock such as sugar and ethanol coupled with ‘low risk’ and rapid catalytic reforming

(BYOGY, Virent). Disadvantages include mass transfer issues such as the slow diffusion of gases into

aqueous fermentation broths and the difficulty of isolating organic acids from fermentation mixtures.

Catalyst issues such as the low tolerance of current reforming catalysts to oxygenated feedstocks are


July 2014 8

also a challenge, as are feedstock and capital intensity to provide hydrogen for catalytic reduction of

acids, alcohols or other oxygenated products to de-oxygenated saturated hydrocarbons.

While tremendous technical progress and commercialization activity have taken place over the past

several years, only relatively small amounts of drop-in biofuels functionally equivalent to petroleum-

derived transportation fuels are commercially available today. In the same way conventional (so-called

“first generation”) bioethanol from sugar and starch was used to establish the infrastructure and “rules”

for subsequent production and use of advanced (so-called “second generation”) bioethanol, it is likely

that oleochemical derived drop-in biofuels will initially be used to establish the markets and procedures

for use of drop-in biofuels. This is exemplified by the many Hydrotreated Vegetable Oil (HVO)-based

biofuel flight trials and refinery processing trials undertaken over the last few years and by the recent

ASTM approval of oleochemical derived jet fuel blendstocks. However, significant expansion of the

oleochemical platform will be limited by the cost, availability and sustainability of food grade (vegetable

oil) or animal oil/fat based feedstocks. The challenge of developing emerging thermochemical based

drop-in technologies can be viewed as analogous to cellulosic ethanol, which uses more plentiful, non-

food lignocellulosic biomass as feedstock but entails larger technology risks and higher capital costs. In

this context, thermochemical technologies are well positioned to account for a considerable component

of drop-in biofuel capacity growth over the near-to-midterm. This is primarily because biochemical and

hybrid based drop-in biofuel processes typically provide lower yields of higher value oxygenated

intermediates (e.g. organic dialcohols and diacids) that can command higher value in the rapidly growing

bio-based chemicals markets. It is also likely that future biorefineries will utilize biomass in much the

same way that current petroleum refineries use crude oil by converting the raw feedstock into a diverse

range of fuels and chemicals products in a single highly integrated facility. However, it is probable that

larger sized thermochemical based facilities will primarily focus on converting biomass feedstocks to

commodity scale drop-in biofuels and bioenergy products while somewhat smaller scale biochemical or

algal platform based facilities will convert sugar, biomass or syngas feedstocks to specific higher value

non-commodity products such as farnesene, butanediol, succinic acid, butanol or oils for use in more

lucrative biobased chemicals markets (e.g., cosmetics, food additives, lubricants, etc.). Regardless, for all

of these technologies, hydrogen sourcing will play a major role in future commercialization of drop-in

biofuel platforms.


July 2014 9

Contents CHAPTER 1: BACKGROUND – BIOFUELS & PETROLEUM INDUSTRY ....................................................... 14

1.1 Biofuels rationale ........................................................................................................................ 14

1.2 Current biofuels .......................................................................................................................... 16

1.2.1 Ethanol ................................................................................................................................ 17

1.2.2 Biodiesel .............................................................................................................................. 21

1.2.3 Biomass-derived biofuels .................................................................................................... 22

1.3 Definition of drop-in biofuels ...................................................................................................... 24

1.4 Reasons for the increasing interest in Drop-in biofuels.............................................................. 25

1.4.1 Blend walls .......................................................................................................................... 25

1.4.2 Energy density, aviation and other long distance transportation sectors .......................... 26

1.4.3 Energy security and crude oil prices ................................................................................... 27

1.4.4 Infrastructure incompatibility ............................................................................................. 27

1.5 Properties of petroleum and drop-in biofuels ............................................................................ 28

1.6 The Oxygen challenge ................................................................................................................. 30

1.7 Deoxygenation of biomass .......................................................................................................... 31

1.7.1 The Hydrogen -Biomass feedstock dilemma (or trade-off) ................................................ 31

1.7.2 The Hydrogen to Carbon ratio ............................................................................................ 32

1.8 Hydrogen in the petroleum Industry .......................................................................................... 34

1.8.1 Declining quality of crude oil .............................................................................................. 34

1.8.2 Demand for low sulfur and light petroleum products ........................................................ 38

1.9 Oil refining basics (emphasis on hydrotreating and hydrocracking) .......................................... 41

1.9.1 Crude oil .............................................................................................................................. 41

1.9.2 Oil refining ........................................................................................................................... 42

1.9.3 Fluid Catalytic Cracking (FCC) .............................................................................................. 43

1.9.4 Hydrotreating ...................................................................................................................... 45

1.9.5 Hydrocracking ..................................................................................................................... 45

1.10 Hydrogen demand in the oil industry ......................................................................................... 46

1.11 Hydrogen generation in oil refineries ......................................................................................... 47

1.12 Hydrogen use in the ammonia fertilizer industry ....................................................................... 49

1.13 Biomass as a source of hydrogen ................................................................................................ 49


July 2014 10

CHAPTER 2: THE OLEOCHEMICAL PLATFORM ........................................................................................ 51

2.2 Process overview ........................................................................................................................ 51

2.2.1 Conventional oleochemical-based biofuel (esterified fatty acids) ..................................... 51

2.2.2 Advanced oleochemical (hydrotreated lipids) .................................................................... 54

2.3 Potential for integration with oil refineries ................................................................................ 57

2.3.1 Challenges of hydroprocessing renewable oils in conventional refineries......................... 57

2.4 Commercialization aspects ......................................................................................................... 60

2.4.1 Feedstock sensitivity ........................................................................................................... 60

2.4.2 Feedstock procurement ...................................................................................................... 61

2.4.3 Commercial facilities ........................................................................................................... 64

2.4.4 Fuel Quality ......................................................................................................................... 65

2.4.5 Test flights and certification................................................................................................ 66

2.4.6 Sustainability certification of HEFA ..................................................................................... 67

CHAPTER 3: THE THERMOCHEMICAL PLATFORM ................................................................................... 69

3.1 Overview of thermochemical processes ..................................................................................... 70

3.2 Fast Pyrolysis ............................................................................................................................... 73

3.2.1 Bio-oil production ............................................................................................................... 73

3.2.2 Bio-oil composition ............................................................................................................. 78

3.2.3 Bio-oil uses .......................................................................................................................... 82

3.2.4 Bio-oil standards ................................................................................................................. 83

3.2.5 Cost of Bio-oil ...................................................................................................................... 85

3.2.6 Hydrothermal liquefaction .................................................................................................. 86

3.3 Upgrading Pyrolysis oils to motor fuels ...................................................................................... 86

3.3.1 Hydrotreating ...................................................................................................................... 87

3.3.2 Zeolite cracking ................................................................................................................... 93

3.3.3 Catalytic pyrolysis ................................................................................................................ 94

3.4 Refinery integration of pyrolysis platforms ................................................................................ 96

3.4.1 Fluid Catalytic Cracking (FCC) co-processing ....................................................................... 98

3.4.2 Co-hydrotreating ................................................................................................................. 99

3.5 Techno-economics and sensitivities of upgrading bio-oils ....................................................... 101

3.6 Pyrolysis biofuel commercialization ......................................................................................... 105


July 2014 11

3.7 Gasification ............................................................................................................................... 108

3.7.1 Gasification process .......................................................................................................... 109

3.7.2 Gasifier types .................................................................................................................... 110

3.7.3 Syngas cleanup .................................................................................................................. 113

3.7.4 Syngas uses ....................................................................................................................... 116

3.8 Syngas catalytic condensation for synthesis ............................................................................. 116

3.8.1 Methanol to Gasoline process .......................................................................................... 117

3.8.2 Fischer-Tropsch process .................................................................................................... 118

3.9 Syngas upgrading techno-economic considerations ................................................................ 123

3.10 Progress in the commercialization of biomass gasification and fuel synthesis ........................ 127

3.10.1 Fixed Bed gasifiers ............................................................................................................. 127

3.10.2 Bubbling Fluidized Bed facilities ........................................................................................ 127

3.10.3 Circulating Fluidized Bed facilities ..................................................................................... 128

3.10.4 Dual Fluid Bed facilities ..................................................................................................... 130

3.10.5 Entrained Flow Gasification facilities ................................................................................ 133

3.10.6 Plasma gasifiers ................................................................................................................. 137

3.10.7 The IEA Bioenergy Task 33 gasification facilities database ............................................... 137

CHAPTER 4: THE BIOCHEMICAL PLATFORM ......................................................................................... 140

4.1 Ethanol ...................................................................................................................................... 141

4.2 Biological conversion to drop-in biofuels ................................................................................. 145

4.2.1 The butanol pathway ........................................................................................................ 146

4.2.2 The fatty acid pathway ...................................................................................................... 149

4.2.3 The isoprenoid pathway ................................................................................................... 151

4.3 Feasibility considerations .......................................................................................................... 152

4.4 Commercialization efforts......................................................................................................... 159

4.4.1 Butanol .............................................................................................................................. 159

4.4.2 Fatty acids, long chain alcohols and alkenes .................................................................... 161

4.5 The value added chemicals opportunity ................................................................................... 162

CHAPTER 5: HYBRID PLATFORMS .......................................................................................................... 167

5.1 Syngas fermentation ................................................................................................................. 168

5.2 The Zeachem process ................................................................................................................ 171


July 2014 12

5.3 Aqueous phase reforming and the Virent process ................................................................... 171

5.4 The alcohol to jet process (ATJ) ................................................................................................ 174

CHAPTER 6: CONCLUSIONS AND RECOMMENDATIONS ....................................................................... 177

Tables

Table 2-1: Properties of HEFA petroleum diesel and FAME biodiesel. ....................................................... 53

Table 2-2: Current world annual production capacity of HEFA drop-in biofuels ........................................ 64

Table 2-3: Select roperties and specifications of fossil and renewable HEFA diesel and jet fuels. ............ 65

Table 3-1: Commercial and pre-commercial (≥ 50 tpd) bio-oil facilities by 2014 ....................................... 78

Table 3-2: Typical properties of wood pyrolysis bio-oil and of heavy fuel oil (HFO) .................................. 80

Table 3-3: ASTM standards for pyrolysis oils .............................................................................................. 84

Table 3-4: Comparison of FCC and hydroprocessing as refinery co-processing platforms for bioils ....... 101

Table 3-5: Recent techno-economic studies on pyrolysis-derived biofuels ............................................. 104

Table 3-6: Char and Tar content of biomass syngas from different reactors ........................................... 114

Table 3-7: Techno-economic studies on gasification-derived biofuels .................................................... 124

Table 3-8: Scale of selected GtL and BtL facilities ..................................................................................... 126

Table 4-1:Scale up issues for anerobic vs aerobic fermentation .............................................................. 157

Table 4-2: Biological process performance metrics for selected biofuel production microorganisms .... 155

Figures

Figure 1-1: Average IEA crude oil price ....................................................................................................... 16

Figure 1-2: Historic biofuel production volumes ........................................................................................ 17

Figure 1-3: Brazilian sugarcane land use and ethanol productivity ............................................................ 20

Figure 1-4: Price fluctuation of food vs cellulosic biofuel feedstocks ......................................................... 23

Figure 1-5: Typical sequence of petroleum products flow through a pipeline ........................................... 28

Figure 1-6: Carbon number and boiling point range of commercial transportation fuels ......................... 29

Figure 1-7: Current commercial biofuels and their oxygen content ........................................................... 30

Figure 1-8: The effect of oxygen content on the energy density of liquid fuels ......................................... 31

Figure 1-9: Simplified representation of carbohydrate deoxygenation mechanisms ................................ 32

Figure 1-10: The effective Hydrogen to Carbon ratio “staircase” .............................................................. 34

Figure 1-11: Purvin & Gertz forecast for world crude oil quality ................................................................ 35

Figure 1-12: The hydrogen to carbon “staircase” for fossil fuel feedstocks ............................................... 38

Figure 1-13: World oil demand by product, 2010-2017 ............................................................................. 39

Figure 1-14: Sulfur Content Specifications for U.S. Petroleum Products, 1990-2014 ................................ 41

Figure 1-15: Simplified diagram of an oil refinery ...................................................................................... 43

Figure 1-16: Simplified representation of a fluidized catalytic cracker ...................................................... 44

Figure 1-17: Simplified depiction of a hydrocracking reaction ................................................................... 46

Figure 1-18: Hydrotreating capacity in the US ........................................................................................... 47


July 2014 13

Figure 1-19: Crude/Natural gas price differentials ..................................................................................... 48

Figure 1-20: Estimated Cost Hydrogen Production from Biomass and Natural Gas Feedstocks ................ 50

Figure 2-1: Triglyceride to Biodiesel (FAME) reaction ................................................................................ 52

Figure 2-2: Simplified Hydroprocessed Esters and Fatty Acids (HEFA) process depicting the 2 stages of

hydroprocessing .......................................................................................................................................... 55

Figure 2-3: Triacyl glyceride (TAG) deoxygenation process ........................................................................ 56

Figure 2-4: Commodity prices of Diesel and Palm Oil ................................................................................. 62

Figure 2-5: Timeline of biofuel test flights and ASTM certification approvals ........................................... 67

Figure 3-1: Product spectrum from thermochemical conversion of biomass ............................................ 71

Figure 3-2: Simplified representation of major thermochemical drop-in biofuel process routes ............. 72

Figure 3-3: Simplified schematic of bubbling fluidized bed (BFB) fast pyrolysis ........................................ 74

Figure 3-4: Simplified schematic of circulated fluid bed (CFB) fast pyrolysis ............................................. 75

Figure 3-5: Pyrolysis oil sample ................................................................................................................... 79

Figure 3-6: Reactivity scale of oxygenated compounds under hydrotreatment conditions. ..................... 90

Figure 3-7: Refinery insertion points (red arrows) for HDO Bio-oils ........................................................... 97

Figure 3-8: Schematic of updraft (a) and downdraft (b) fixed bed gasifiers ............................................. 111

Figure 3-9: Schematic of a biomass pretreatment via fast pyrolysis followed by an entrained-flow gasifier

.................................................................................................................................................................. 113

Figure 3-10: Main Syngas Conversion Pathways ...................................................................................... 117

Figure 3-11: Simplified representation of Fischer-Tropsch chain growth on a catalyst surface .............. 119

Figure 3-12: Simplified schematic of the overall biomass gasification to FT diesel and gasoline process122

Figure 3-13: The Goteborg biomass gasification to methane facility ....................................................... 131

Figure 3-14: The Choren gasification/FT process ...................................................................................... 134

Figure 3-15: The IEA Bioenergy Task 33 gasification facility database ..................................................... 138

Figure 4-1: Schematic overview of the microorganism-to-drop-in-intermediate or biofuel derived for

various renewable feedstocks .................................................................................................................. 141

Figure 4-2: Simplified schematic of major metabolic pathways relevant to drop-in biofuel production 146

Figure 4-3: Biosynthesis of fatty acid, palmitate ...................................................................................... 150

Figure 4-4: example isoprenoids ............................................................................................................... 152

Figure 4-5: The H/Ceff ratio in relation to the biochemical platform and its intermediates .................... 163

Figure 5-1: The Virent process. (a) Entire process. (b) simplified representation emphasizing key steps.

.................................................................................................................................................................. 173


July 2014 14

CHAPTER 1: BACKGROUND – BIOFUELS & PETROLEUM INDUSTRY

Biofuels are currently being developed as renewable alternatives to fossil derived transportation fuels

with the hope of achieving environmental and socioeconomic benefits such as reduced GHG emissions,

employment generation and energy security. Bioethanol and biodiesel are the main commercially

available biofuels and currently contribute about ~2% by volume of global transportation fuel demand

(US EIA, 2013). However, these fuels are chemically and functionally different from petroleum-derived

fuels and they thus do not make full use of the existing petroleum processing and distribution

infrastructure. As infrastructure components, such as vehicle engines, fueling stations, refineries, etc.,

are very expensive to change, it is recognised that it would simplify biofuels production and usage

growth if biofuels could be readily “dropped-into” the existing infrastructure (petroleum distribution

and refining, fuel specifications, etc.) and be functionally equivalent to current petroleum-derived fuels.

However, as will be described in more detail, it is likely that producing such “drop-in” biofuels will

require more complex processing infrastructure and higher processing inputs, most notably hydrogen

(H2) inputs, than today’s predominant bioethanol and biodiesel biofuels. Consequently, greater techno-

economic challenges will probably be encountered when trying to achieve cost competitive routes to

drop-in biofuels

Hydrogen is a key input not only for drop-in biofuel producers (e.g. hydrotreated vegetable oils) but also

for other sectors, most notably the Oil & Gas sectors, which have to upgrade crude oil of ever declining

quality to meet the needs of a growing market for more refined and lighter petroleum products. In the

future, drop-in biofuel producers may have to compete for hydrogen resources with the petroleum

sector as well as the ammonia fertilizer industry.

1.1 Biofuels rationale

Unstable and rising petroleum prices, the finite nature of the resource, as well as concerns about fossil

fuel emissions and dependence on politically unstable regions for transportation fuel imports are among

the major motivations for pursuing biofuels. Biofuels are arguably the most likely near term renewable

alternative to petroleum fuels, with some forms of transportation that cannot be easily electrified (such

as long distance trucking, shipping and aviation) having this approach as the only alternative.

http://www.eia.gov/forecasts/steo/report/global_oil.cfm


July 2014 15

Rising oil prices have been a major motivator in finding alternatives to fossil based transportation fuels,

with the OPEC oil crises of the 1970’s stimulating several counties to look into alternative, renewable

biofuels. However, as oil prices decreased in the 1980’s (Figure 1-1), the interest and research support

into biofuels largely decreased, although countries such as Brazil and the United States continued to try

to commercialise conventional (so-called “first generation”) biofuels, primarily bioethanol from sugars

and starches (grains) and biodiesel from plant crop seed oils. Investment in advanced (so-called “second

generation”) biofuels produced from non-food feedstocks increased in the early 2000s due to a

combination of factors such as increasing awareness of the role of transport derived carbon emissions in

climate change and increasing dependence on crude oil imports to Europe and North America. The

availability and price of these crude oil imports is increasingly uncertain for a variety of reasons such as

geopolitical conflict and political uncertainty. More recently, despite the discovery of new

unconventional Oil & Gas resources, the IEA forecasts that the price of oil will remain high over the

coming decades while demand for biofuels is expected to further increase and play a major role in

meeting ambitious GHG emission reduction targets globally. According to the most likely future policy

scenarios (“current” and “new” policies scenarios) described in IEA’s 2012 Word Energy Outlook (Figure

1-1), the price of oil is expected to rise to above USD $120/barrel by 2035, and in the absence of new

policy action it is projected to climb to almost USD $150/barrel by 2035 (IEA, 2012b).

In this backdrop of expensive petroleum through to 2035, biofuels are well positioned to become a

significantly larger contributor to the global energy landscape. According to the IEA’s “blue map”

scenario, biofuels could provide 27% of total transport fuel by 2050 (IEA, 2011b). If this production level

of biofuel could be achieved it would avoid the production of 2.1 gigatonnes of CO2 emissions per year

compared to if this amount of petroleum-derived fuels were used (IEA, 2011b).


July 2014 16

Figure 1-1: Average IEA crude oil price (Source: IEA, 2012)

1.2 Current biofuels

Conventional (or what others have termed “first generation”) food crop-based biofuels such as

bioethanol from sugar or starch and biodiesel from oilseed, waste oil or tallow are currently the only

commercially available large-scale biofuels. Of the 89 million barrels per day (mbpd) of liquid fuels

produced globally in 2011, 1.9 mbpd were conventional biofuels (US EIA, 2012). Biofuel production has

grown almost exponentially over the last decade (Figure 1-2) with bioethanol providing the vast majority

of this biofuel which is predominantly produced in the USA and Brazil. Recent IEA estimates project

global biofuel production to more than double by 2035 (to ca. 4.5 mbpd) over 2011 production levels

(IEA, 2012b). The majority of the growth in biofuels over the last decade has come from the US and to a

lesser extent Brazil. Together Brazil and the USA currently account for three quarters of world total

biofuel production and close to 90% of global bioethanol production (U.S. EIA, 2010). The world’s third

biggest biofuel producer is the European Union. In contrast to the US and Brazil, the EU produces mostly

biodiesel (>80% of total biofuel production volume) with lesser amounts of bioethanol.

The historical policy, market and infrastructure parameters that made these countries leaders in biofuel

development are relevant to current efforts to develop “drop-in” biofuels.


July 2014 17

Figure 1-2: Historic biofuel production volumes (Source: data from US EIA, 2012)

1.2.1 Ethanol

In 2011 the world produced 1,493,000 bpd of bioethanol fuel with the US producing 908,000 bpd (or

60% of global), Brazil 392,000 bpd (or 25% of global) and the EU 72,000 bpd (or 5% of global) (US EIA,

2011).

The USA is currently the world’s largest ethanol producer. Significant production and use of bioethanol

began in the USA in the early 1980s with the main driver being energy security concerns arising from the

rapid increases in global petroleum prices during the 1970s and 1980s (Tyner, 2008). Another driver that

promoted corn ethanol in the USA was (and still is) the strong campaigns aimed at gaining political

support for expanding markets and revenues for the corn and bioethanol industries. Federal as well

state government measures such as direct funding of partnerships, research funds, tax incentives and

renewable fuel mandates were developed to help the then emerging corn ethanol industry (Mabee,

2007). These US policies, variations of which are still in force, were successful in developing the rapid

increase in ethanol production over the last few decades. Leveraging on an efficient and highly

0

200

400

600

800

1000

1200

1400

1600

1800

2000

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

mill

ion

bar

rels

pe

r d

ay

Rest of the world

EU-27

Brazil

USA

http://www.eia.gov/cfapps/ipdbproject/iedindex3.cfm?tid=79&pid=80&aid=1&cid=CG1,&syid=2007&eyid=2011&unit=TBPD


http://www.agecon.purdue.edu/news/financial/Bioscience_2008_Tyner.pdf


July 2014 18

productive corn industry especially in the US Midwest where corn growing is favoured by both

agronomic and geo-climatic factors, these ethanol supporting policies helped the US become the current

world leader in bioethanol production volume, surpassing Brazil beginning in 2004.

Current US biofuel policies are derived from the Energy Independence and Security Act of 2007 and the

associated Renewable Fuel Standard (RFS). In July 2012 the updated RFS known as “RFS2” came into

effect and stipulated an aggregate of 36 billion gallons (136 billion L) per year of renewable

transportation fuel to be used by 2022. Conventional biofuel (essentially corn grain derived ethanol) is

expected to provide a large portion of this mandate, reaching a plateau of 15 billion gallons (57 million

L) per year by 2015. Advanced biofuels (biomass derived diesel, cellulosic biofuels and non-cellulosic

advanced biofuels) are expected to make up the balance by providing, by 2022, a total of 21 b gallons of

the 36 b gallons per year, with 16 b of these advanced biofuel gallons derived from cellulosic feedstocks.

RFS2 requires biofuels to achieve minimum life-cycle GHG emission reductions relative to petroleum

fuels. If the biofuel LCA demonstrates less than 20% GHG emission reduction compared to the fossil fuel

it displaces it will not be eligible to qualify as a contribution towards the RFS mandate obligations. If the

LCA shows a GHG reduction between 20% and 49%, the biofuel is eligible to count towards the 15 b

gallon conventional fuel obligation (However, this “mandate” is already close to saturated by the

existing corn ethanol industry). Finally, if the biofuel LCA indicates a relative GHG savings of 50% or

higher, it is considered advanced biofuel counting towards the 21 b gallon mandate. Although the RFS’s

annually increasing advanced biofuels volume targets are currently behind schedule, there are ongoing

efforts to achieve the ultimate 21 billion gallons per year advanced biofuel goals by 2022 (Schnepf &

Yacobucci, 2013). Interestingly, Brazilian sugarcane ethanol and, more recently, sweet sorghum ethanol

(grown and distilled in the US using renewable power) have been assessed and qualified to be classified

as advanced biofuels, despite the fact that they are “food” based. This is primarily due to their more

favorable life cycle GHG emission profiles (Schnepf & Yacobucci, 2013).

Excise taxes in the form of VEETC (Volumetric Ethanol Excise Tax Credit) were enacted in 2004 by the

“American Jobs Creation Act”. Under this scheme, blenders can claim tax breaks of $0.5 USD/gal ($0.13

USD/L) of ethanol and $1 USD/gal ($0.26 USD/L) of biodiesel (0.5 USD/gal if made from waste oils).

These tax credits represented a major source of financial support for the growth of the US biofuels

industry. However, these have expired at the end of 2012.


July 2014 19

Pure ethanol cannot be used as unblended fuel in most current automobile vehicle engines without

modifications. One exception is Brazil, where, as will be discussed in the next paragraph, Flexi-fuel

Vehicles (FFVs) which can use neat hydrous ethanol (~E95) are abundant. Another exception are the E85

FFVs that have been designed to use up to 85% by volume anhydrous ethanol. A relatively small number

of E85 vehicles are currently on the road, mostly in the EU and the US. As an example, about 7 million

E85s vehicles were in use in the US in 2012, out of a total light duty vehicle fleet of more than 150

million. The use of ethanol at blend rates beyond 10% may also pose infrastructure compatibility issues

with the rest of the petroleum distribution and processing network. More detail on this incompatibility

of ethanol fuels is provided in Section 1.4.1. The bottom line is that ethanol is not viewed as a fully

“infrastructure compatible” fuel in countries other than Brazil. In the US, for example, blending of

ethanol with gasoline is usually limited to 10% by volume (E10) for regular non-FFVs (flexible fuel

vehicles) and this blending limit is regulated by the EPA (US EPA, 2013). As this blending limit effectively

caps ethanol consumption, it has also been called the “blend wall”. Although the EPA has recently

permitted the use of 15% ethanol by volume for vehicles manufactured in year 2001 or later, the entire

light duty vehicle fleet is not approved to use this higher blend.

Brazil, the second largest ethanol producer, has the world’s oldest and most established biofuel

program. In 1975, triggered by the 1970’s OPEC oil crisis, the Brazilian military led a government

initiated campaign named “National Alcohol Program Proalcool” to make Brazil independent of foreign

oil imports and provide a stable internal demand for its growing sugarcane industry. The program

involved subsidies and incentives for sugarcane ethanol while the government also signed agreements

with automobile manufacturers to help them create a market for vehicles running on pure (hydrous)

ethanol. The campaign was initially so successful that a decade later, in 1985, 100% of new cars sold ran

only on ethanol fuel. However, a few years later due to sugar prices increasing and crude oil prices

decreasing the Brazilian bioethanol industry entered a challenging period. Between the late 1980s and

1999 when the price of hydrated alcohol was not regulated, ethanol use and ethanol-powered car sales

decreased substantially. In response the government introduced ethanol blending mandates (22-25%)

and deregulatory decrees to try to keep the ethanol industry afloat. As petroleum prices increased

during the 2003-2008 period and as E100 FFVs were gradually introduced, ethanol has established itself

as a major component of the Brazilian transportation fuel infrastructure. The success of the Proalcool

program in Brazil is evident in that the sugarcane industry now accounts for 3.5% of GDP and 3.6 million

jobs, with ethanol production consuming about 50% of Brazil’s total sugar feedstock (This latter


July 2014 20

percentage varies depending on the relative commodity prices of sugar and ethanol) (de Almeida et al.,

2008).

Today there are no direct subsidies for ethanol production in Brazil, though there is a degree of

preferential treatment for ethanol as compared to gasoline. Ethanol faces no excise tax while federal

duties are much lower than those for gasoline ($0.26 vs $0.01 per litre). State enforced fuel VAT is also

lower for ethanol than gasoline in most ethanol producing states such as São Paolo. De Almeida et al.

(2008) estimate that ethanol enjoys tax incentives of about USD 1 billion per year and that the Proalcool

program cost around 16 billion for the period of 1979 up until the mid-1990’s (these numbers are much

lower than historical US government support for corn ethanol ) (Sorda et al., 2010). In 2012, ethanol

accounted for 50% market share of the gasoline-powered vehicle fleet in Brazil, with E25 mandatory

across the country and hydrous ethanol (~E95) widely available too.

Figure 1-3: Brazilian sugarcane land use and ethanol productivity (Source: Sawaya Jank, 2011)


July 2014 21

The Brazilian sugarcane industry is well positioned to expand without clearing forest land or displacing

pastures (Figure 1-3) according to UNICA (Brazilian Association of Sugarcane Producers). The current

average yield of ethanol (about 7000 L/ha, as shown in Figure 1-3) is expected to continue to increase.

As depicted in Figure 1-3, out of the 338 m ha of arable land, 30% is unutilized and potentially available

for sugarcane plantations expansion. Sugar cane production currently occupies only 2.6% of Brazil’s

arable land.

The EU accounts for about 5% of global ethanol production using mainly grain starch and beet sugar as

feedstocks. The EU has a strong policy push for biofuels as discussed in section 1.2.2. During the summer

of 2013, the capacity of EU ethanol production increased with the commissioning of Vivergo Fuels Ltd.’s

large wheat-to-ethanol plant in Hull, United Kingdom. This 420 MMly (110 MMgy) facility cost about

$450 million to build and is a joint venture between AB Sugar, BP and DuPont Industrial Biosciences.

The remaining 10% of global ethanol volume production occurs in countries outside the EU, US and

Brazil including, China (39,000 bpd in 2011), Thailand (9,000 bpd in 2011), Australia (7,500 bpd in 2011),

and other countries mainly in Asia and Oceania (US EIA, 2011).

1.2.2 Biodiesel

The EU is the world’s leading biodiesel producer. Of the 0.4 mbpd of biodiesel produced globally in

2011, around 0.2 million were produced in the EU (US EIA, 2012). The US, Brazil and Argentina follow

the EU and each produce about 0.05-0.06 mbpd of biodiesel (US EIA, 2012). Following Rudolf Diesel's

model engine which ran for the first time in Augsburg, Germany in 1893 using biodiesel as a fuel,

biodiesel has been the predominant biofuel used in Europe’s transportation sector, primarily because of

the high proportion of diesel cars in Europe. Other factors that have also contributed to this European

preference are that vegetable oil and animal fats (the biodiesel raw materials) are more regionally

available than starch or sugar (the bioethanol raw materials) coupled with the ease of scaling down

oilseed presses and methesterification units, which makes biodiesel production relatively well suited to

the more decentralized and small scale nature of European agriculture.

Biodiesel is typically comprised of fatty acid methyl esters (FAMEs) derived from vegetable oils (and

animal fats) which, like starch and sugar, compete with food and feed markets. Biodiesel use is typically

limited by blend limits (maximum of 7% biodiesel blend with petroleum diesel). Use of pure biodiesel

http://www.ethanolproducer.com/articles/10022/uk-wheat-to-ethanol-plant-officially-opens-for-business


http://www.eia.gov/cfapps/ipdbproject/iedindex3.cfm?tid=79&pid=81&aid=1&cid=regions,&syid=2011&eyid=2011&unit=TBPD

http://www.eia.gov/cfapps/ipdbproject/iedindex3.cfm?tid=79&pid=81&aid=1&cid=regions,&syid=2011&eyid=2011&unit=TBPD


July 2014 22

(B100) requires engine modifications to avoid maintenance and cold flow performance problems

(Knothe, 2011).

Europe’s biodiesel and bioethanol policies are primarily derived from the EU Renewable Energy

Directive (RED, 2009/28/EC). In April 2009, the European parliament endorsed the RED and a binding

target of 10% renewable energy use in transport (mainly to be met by biofuels) in 2020. In the same

directive, the minimum GHG saving requirements were specified (35% initially and 50% starting from

2017) as well as other sustainability criteria such as no feedstock from protected habitats and other

territories of ecological value. Biofuel tax incentives or mandated biofuel market shares were to be set

by national governments and they differ from one European country to another. Recently the EU

Parliament voted for a 6% by volume cap on food-derived biofuels while setting a separate 2.5% target

to incentivize non-food biofuels, made from waste products and lignocellulosic biomass. The inclusion of

indirect land use (iLUC) factors in accounting for a fuel’s carbon footprint post-2020 has also been

proposed by the EU Parliament (van Noorden, 2013). This 6% cap and the inclusion of iLUC factors have

however not been agreed upon between the EU Parliament and the Council (EU member state

governments). A decision on these biofuel policies has created a “deadlock” over the last year or so and

negotiations were still continuing as of 30 May 2014 (Euractiv, 2014).

1.2.3 Biomass-derived biofuels

The use of food products such as sugar and vegetable oils as raw materials for the manufacturing of

biofuels has raised concerns about this increased use detrimentally driving higher prices for food/feed.

The use of woody or grassy fibers (lignocellulosic biomass) is viewed as a more acceptable route for

further growing the production of renewable liquid fuels. Such lignocellulosic materials (“biomass”) are

more abundant and have good potential to provide higher fuel yields per unit of land area than food

crops. With continued development, conversion of biomass feedstocks can be made to be less costly

and to achieve lower carbon emission intensity than oilseed, starch or sugar crops. Historic price

fluctuations have been much higher for food feedstocks than for lignocellulosic feedstocks. For example,

although sugar and palm oil prices have fluctuated considerably since 1995 (Figure 1-4), over this same

time period the price for biomass, such as hardwood logs, has been much more stable.

The main fibrous biomass feedstocks being considered for biofuels are crop residues (corn stover, wheat

straw, sugarcane bagasse, etc.), herbaceous energy crops (miscanthus, switchgrass, cardoon, etc.), wood


July 2014 23

derived materials such as sawmill and forest residues and fast-rotation forestry species such as polar

and willow.

Figure 1-4: Price fluctuation of food vs cellulosic biofuel feedstocks (Data from indexmundi 2013: http://www.indexmundi.com/commodities/)

The conversion of fibrous biomass (lignocellulose) to liquid fuels is achieved by two main process

pathways, thermochemical and biochemical. Thermochemical processes aim at converting the bulky

solid biomass to an energy dense liquid using combinations of pressure, temperature and catalysts

during the conversion processes. Biochemical processes aim at biologically converting the biomass first

to sugars then to a liquid fuel molecule such as ethanol. This latter type of process typically involves the

integrated process steps of pretreatment, enzymatic hydrolysis, biological conversion (e.g.,

fermentation) and concentration. The pretreatment step (chemical and/or physical) allows fractionation

of the lignin, hemicellulose and cellulose, and increases the accessibility of the cellulose to hydrolyzing

agents such as enzymes. The sugars in the resulting enzymatic hydrolyzate are then fermented to

ethanol or butanol or biologically converted to other liquid fuel molecules (e.g., farnesene), which are

then separated (recovered) by distillation, liquid-liquid separation, membranes or other means.

0

5

10

15

20

25

30

35

0

200

400

600

800

1000

1200

1400Ja

n-9

5

Jan

-96

Jan

-97

Jan

-98

Jan

-99

Jan

-00

Jan

-01

Jan

-02

Jan

-03

Jan

-04

Jan

-05

Jan

-06

Jan

-07

Jan

-08

Jan

-09

Jan

-10

Jan

-11

Jan

-12

US

cen

ts

US

Do

llars

Hard LogsPrice (USDollars perCubicMeter)

palm oil(US dollarsper metrictonne)

Sugar Price(US centsper Pound)

http://www.indexmundi.com/commodities/


July 2014 24

However, while many thermochemical and biochemical routes to cellulosic biofuels are available, the

processing of fibrous feedstocks to fuel grade products requires more energy and resources than does

conventional biofuels (Sims et al., 2008).

As is discussed in more detail below, “drop-in” biofuels are expected to demonstrate the same

functional properties as petroleum-derived fuels. These types of “drop-in” biofuels can be produced

from cellulosic biomass via thermochemical conversion, from lipid feedstocks via hydrotreatment, or

from sugars and alcohols via biological or chemical catalysis. Examples of such biofuels include Fischer-

Tropsch liquids (FT liquids), hydrotreated pyrolysis oils (HPOs), and hydrotreated vegetable oils (HVOs).

More recently, biochemical pathways, such as sugar conversion to drop-in or close to drop-in

hydrocarbon molecules have also been developed and proposed as candidate technologies for

hydrocarbon biofuel production. Companies such as Amyris and LS9 have engineered microorganisms to

convert sugar to “diesel-like” molecules. Currently, these drop-in biofuel processes are largely at the

research and demonstration stages although several large scale HVO facilities are being built and

operated by companies such as Neste Oil and Dynamic Fuels while a commercial scale pyrolysis and

upgrading facility was recently built and commissioned by the company KiOR.

1.3 Definition of drop-in biofuels

Conventional biofuels have a distinct chemical nature and so they can be accurately defined by their

chemical composition alone. For example, bioethanol is ethanol and biodiesel is a fatty acid methyl ester

(FAME). In contrast drop-in biofuels generally consist of a mixture of many different types of

hydrocarbons, the properties of which, just like petroleum fuels, is typically characterized by the

mixtures’ functional characteristics such as distillation profile, viscosity, acidity, etc. A true drop-in

biofuel should be able to be readily “dropped in” to the existing petroleum infrastructure and be

handled in much the same way as petroleum fuels without requiring significant infrastructure

adjustments.

In this vein, and considering the diversity of drop-in biofuel processes and product options, the following

definition is used throughout this report to provide a functional representation of what is meant by a

drop-in biofuel:


July 2014 25

“Drop-in biofuels are liquid bio-hydrocarbons that are functionally equivalent to petroleum fuels and are

fully compatible with existing petroleum infrastructure”

It should be noted that petroleum itself can sometimes contain up to 2 wt% oxygen (infrequently, even

more) (Speight, 2006). The term “petroleum-derived blendstocks (fuels)” is used to describe the gasoline

(petrol), diesel, jet and other types of commercial transportation fuel blendstocks as well as their

refinery precursors that are currently processed in existing refineries, pipelines or anywhere upstream

of a blending terminal in the petroleum supply chain.

1.4 Reasons for the increasing interest in Drop-in biofuels

Drop-in biofuels are currently attracting considerable attention. Some of the reasons are directly or

indirectly related to challenges to further increasing the markets for ethanol and biodiesel biofuels such

as their likely blend wall and supply constraints. Drop-in biofuels are better positioned as they avoid

blend wall concerns and also potentially make better use of existing infrastructure (current inventory of

petroleum refineries, supply channels and liquid fuel powered combustion engines).

1.4.1 Blend walls

As mentioned earlier, bioethanol and biodiesel cannot be used in a neat form in conventional

automobile engines (without modifications and tuning) and they are not fully fungible with existing

petroleum fuels. As a result, there are limits on the blending levels of these biofuels with petroleum

fuels, with these limits stipulated and regulated by governments after consultation with automobile

manufacturers and oil companies. With the exception of Brazil’s FFVs and US and EU E85 vehicles, most

jurisdictions outside of Brazil blend ethanol at levels that do not exceed 10% by volume (E10). The

blending rate for biodiesel generally varies between 2% and 20% by volume. This blend wall has, in part,

limited the growth of biofuels and in the US in particular, these ethanol blend wall volumes have already

been reached (Tyner, 2010). To confuse matters further, the RFS ethanol mandate currently stipulates

consumption of ethanol volumes at levels above what can be used without breaking the E10 blend wall.

Short of buying ethanol that they cannot sell, gasoline blenders have resorted to buying the ethanol

permit equivalent (Renewable Identification Numbers) from non-obligated parties. This RIN trading has

given rise to an unregulated futures market which resulted in the recent, unprecedented surge in the


July 2014 26

market price of US RFS’s ethanol RINs from under USD $0.10/gallon ($0.026/L) to over US $1/gallon

($0.26/L). This has triggered increased tensions between RFS-obligated oil companies and the US

Environmental Protection Agency (Schnepf & Yacobucci, 2013). Although one solution to resolve the

blend wall constraint would be the wider use of E85 flexi-fuel vehicles (FFVs), as shown by Tyner and

Viteri (2010), E85 penetration cannot grow fast enough to provide a fleet that could absorb all the

“over-the-blend wall” amounts of ethanol that could be produced in the US. If there was enough

economical drop-in biofuel production in the US, such blend wall-related issues would be avoided.

1.4.2 Energy density, aviation and other long distance transportation sectors

Aside from blend wall limitations in gasoline and diesel automobiles, there are transportation modes

where conventional biofuels cannot be used, or their use is not favoured. Aviation is the most salient

example of a transportation sector that can only use drop-in biofuels since ethanol and biodiesel do not

fulfill key jet fuel requirements such as stringent cold flow viscosity and energy density specifications.

Since jet engines cannot be readily “electrified” they are uniquely dependent on biofuels for renewable

fuel alternatives. In addition, the aviation industry’s requirements for affordable and renewable jet fuels

are becoming ever more pressing, as the industry has committed to GHG emission reductions amidst

increasing oil prices and increasing demand for air travel. Examples of such aviation biofuel

commitments include (not an exhaustive list):

European Commission’s "European Advanced Biofuels Flight path" initiative is a roadmap with

clear milestones to achieve an annual production of two million tonnes of sustainably produced

biofuel for aviation by 2020.

US Federal Aviation Administration’s (FAA) aviation biofuel goal to use 1 billion gallons (3.8

million L) of renewable jet fuel per year from 2018 onwards (Hileman et al., 2009)

Various non-binding strategic targets for reducing aviation’s carbon footprint via private airline

carriers such as Lufthansa and Alaska Airlines as well as international aviation alliances such as

the IATA (IATA, Dec 2013) and SAFUG (SAFUG, Dec 2013).

KLM Airlines in collaboration with Air France, Argos and Spring Associates recently (2009)

formed SkyNRG, a major international broker of available bio-jet fuels.

Regional organisations/consortia for the development of aviation biofuels such as the

Sustainable Aviation Fuels Northwest (SAFN) in the US and the Brazilian coalition of academic,

government and commercial partners: Aliança Brasileira para Biocombustiveis de Aviação

(ABRABA).

http://ec.europa.eu/energy/renewables/biofuels/flight_path_en.htm

http://www.lufthansagroup.com/en/responsibility/climate-and-environmental-responsibility/keroseneandemissions/biofuel-at-lufthansa.html

http://www.alaskaair.com/content/about-us/social-responsibility/fly-green/about-sustainable-aviation-biofuels.aspx

https://www.iata.org/pressroom/facts_figures/fact_sheets/pages/alt-fuels.aspx

http://www.safug.org/

http://skynrg.com/our-story/


July 2014 27

More details on drivers behind the need for developing aviation biofuels are described in a recent IEA

Bioenergy Task 40 report (Rosillo-Calle et al., 2012) as well as a report by the Air Transport Action Group

(ATAG, 2012), a report by the Sustainable Aviation Fuels Northwest initiative (SAFN, 2011) and the

Aviation biofuels website of the European Biofuels Technology Platform. Other long distance and non-

electrifiable transportation modes such as marine shipping and long distance trucking are also better

suited to using drop-in biofuels than conventional biofuels. It should be noted that these sectors are

expected to represent much of the transportation fuel demand growth over the next decade or so (IEA,

2012). As discussed in more detail in Section 1.8.2, it is expected that regulations will be tightened on

maximum sulfur content allowed in marine and road fuels. Consequently, biomass derived drop-in

biofuels (and other biofuels) that exhibit low sulfur content will also look more attractive.

1.4.3 Energy security and crude oil prices

As explained in the IEA’s “blue map” scenario (IEA, 2011b) biofuels are projected to account for 27% of

total global transportation fuel demand by 2050. This ambitious target will be difficult to reach using

only conventional biofuels such as ethanol and biodiesel.

1.4.4 Infrastructure incompatibility

Due to the chemical nature of bioethanol and biodiesel they have to be delivered and blended through

separate distribution channels as they are incompatible with much of the petroleum infrastructure such

as pipelines and storage tanks. Thus alternative channels must be used such as truck, rail or barge and

this adds to the cost and carbon footprint of biofuels. The majority of the petroleum distribution

infrastructure, such as pipelines, tanks, and related equipment is composed of low carbon and low alloy

steels, and controlling rust and corrosion is of primary importance. Pipelines run petroleum products in

batches which follow one after the other as shown in Figure 1-5. These batches follow specific

sequences in order to avoid cross contamination. Between batches, a small amount of co-mingled

product, known as interface or transmix, is generated and is normally segregated for refractionation to

diesel and gasoline or returned to a refinery for processing. In this regard pipelines are vulnerable to

contamination which can carry over from batch-to-batch. Biodiesel, for example, is reactive with

pipeline metallurgy and it can adhere to the surfaces of pipeline walls, potentially contaminating

subsequent petroleum batches. Jet fuels are particularly sensitive to biodiesel ester contaminants. As a

general rule, any potential biofuel that might be transported through petroleum pipelines must be non-

corrosive and hydrophobic. Most pipeline networks have engineering features in place to remove

http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:The+Potential+and+Role+of+Biofuels+in+Commercial+Air+Transport+-+Biojetfuel#0

http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:The+Potential+and+Role+of+Biofuels+in+Commercial+Air+Transport+-+Biojetfuel#0

http://www.atag.org/component/downloads/downloads/152.html

http://www.safnw.com/sustainable-aviations-fuels-bibliography/

http://www.biofuelstp.eu/air.html


July 2014 28

contaminating water (Bunting et al., 2010). An ideal drop-in biofuel would have similar (to petroleum)

non-corrosive, non-reactive and non-hydrophilic functional properties so that it can fully utilize the

existing, substantial, pipeline network for its distribution.

Figure 1-5: Typical sequence of petroleum products flow through a pipeline (Source: adapted from API, 2001)

1.5 Properties of petroleum and drop-in biofuels

As defined earlier, drop-in biofuels must be functionally equivalent to current gasoline, diesel, jet and

related fossil derived transportation fuels. This functional equivalence implies that drop-in biofuels must

meet certain bulk properties such as miscibility with petroleum fuels, compatibility with fuel

performance specifications, good storability, transportability with existing logistics structures and

usability within existing engines (vehicles, jet planes etc.). Chemically, a drop-in biofuel could also be

defined as a biomass-derived liquid hydrocarbon that has a low oxygen content, low water solubility and

a high degree of carbon bond saturation. The exact specifications of these fuels will be determined by

various physicochemical properties such as viscosity, carbon number, boiling point range, freezing point,

flash point, aromatic content and others. Of these various properties the carbon number and boiling

point range (shown in Figure 1-6) are the most commonly used parameters to distinguish between

gasoline (light distillate), diesel and jet fuels (middle distillates). Gasoline is typically used in spark

ignition engines and comprises a mixture of C4-C12 hydrocarbons with a 20-40% aromatic content.

Diesel is primarily used in compression engines and it contains C10-C22 hydrocarbons with a 25%

aromatic content. Aviation fuel is a mixture of C8-C16 with a maximum of 25% aromatic content and a

range of stringent specifications such as very low freezing point (-40 C), thermal stability and low

DIESEL fuel

Regular gasoline

Premium gasoline

Regular gasoline

DIESEL fuel

Regular gasoline

TRANSMIX (interface material which must be reprocessed)


July 2014 29

viscosity at low temperatures (Hileman et al., 2009). Marine fuel is a lower quality and lower cost fuel

that is derived from the heavier distillates of refineries which contain very long carbon chains and low or

no phenolics. The quality of marine fuels is measured with viscosity and density indexes (similarly to

crude oil) rather than boiling point ranges (Vermeire, 2012).

Regarding biofuel properties and as shown in Figure 1-6, ethanol falls within the boiling point and

carbon number range of gasoline fuels but it is only partly blendable with gasoline without the need for

engine modification such as in flexifuel vehicles. Similarly biodiesel fits the properties of diesel fuel but it

is again only partly blendable. As far as jet fuel properties are concerned, neither ethanol nor biodiesel

fall within the narrow carbon number range of jet fuels and thus (as stated earlier) these fuels are not

suited for aviation. While there are about a dozen other properties that have to be met by fuels in order

to qualify for ASTM certification1 as usable in existing engines, the oxygen content (or H/C ratio) and

carbon number range can be considered as the minimum, most basic characteristics that have to be

met.

Figure 1-6: Carbon number and boiling point range of commercial transportation fuels Source: (Hileman, Ortiz, Bartis, & Wong, 2009)

1 For #2 Diesel Fuel for example, the ASTM standard contains the following parameters: Particulate Contamination by Filtration, BP Distribution, Ash, Carbon, Hydrogen and Nitrogen, Carbon Residue, Cloud Point, Acid and Base Number, Color, Cold Filter Plugging Point, Copper Corrosion, Density, Distillation, Flash Point, Heat Content, Hydrocarbon Type, Lubricity (HFRR), Pour Point, Sulfur Content, Vicosity, Kinematic, Water and Sediment, High Temp Stability (Source: www.astm.org)

0

50

100

150

200

250

300

350

400

0

2

4

6

8

10

12

14

16

18

20

22

24

Ethanol Gasoline Jet A Diesel Biodiesel

bo

ilin

g p

oin

t (°

C)

carb

on

nu

mb

er

carbon number

boiling point


July 2014 30

1.6 The Oxygen challenge

The greatest challenge for drop-in biofuels to meet the physicochemical properties of petroleum-based

transportation fuels is decreasing the high oxygen content of biomass derived biofuels. Oxygen is

present in biomass in various chemical functional groups such as esters, ethers and hydroxyl groups.

While this oxygen content is potentially valuable for metabolic processes and for the production of

some value added chemicals, it is highly undesirable for drop-in biofuels. As shown in Figure 1-7,

biodiesel and bioethanol are only partially deoxygenated and this is one of the main reasons why these

conventional biofuels are not fully compatible with existing petroleum infrastructure. These oxygenated

functional groups can react with refinery and pipeline metallurgy as well as with biofuel components to

form gums acids and other impurities often at the detriment of biofuel storability/stability (Pearlson,

2011; Bridgwater, 2012).

Figure 1-7: Current commercial biofuels and their oxygen content

Compatibility and reactivity are not the only reasons why it is important to deoxygenate biofuels. The

oxygen in biofuels reduces their energy density. This in turn determines the size of a vehicle’s fuel tank

which in turn determines travel range for all modes of transportation. As shown in Figure 1-8, with

increasing oxygen content, expressed as the molar ratio of oxygen to carbon (O/C) in the fuel molecule,

the energy density of biofuels and biomass processing intermediates decreases linearly.

Ethanol Biodiesel (fatty acid methyl ester)

CH3


July 2014 31

Figure 1-8: The effect of oxygen content on the energy density of liquid fuels Data from ORNL, 2013

1.7 Deoxygenation of biomass

1.7.1 The Hydrogen -Biomass feedstock dilemma (or trade-off)

Deoxygenation of biomass intermediates is essential for the production of drop-in biofuels and it is

primarily achieved by two main chemical reduction processes, hydrodeoxygenation (HDO) and

decarboxylation (DCO). During hydrodeoxygenation the hydrogen present in the biomass intermediates

(or supplied externally) is oxidized and oxygen is removed as water (H2O) while in decarboxylation the

carboxyl group carbon is oxidized and the oxygen is removed as carbon dioxide. While in practice these

two processes take place simultaneously, certain conditions favour one reaction over the other. The

HDO process is typically favoured when hydrogen can be readily accessed from an external source (e.g.

hydrogen gas derived from natural gas) while, in the absence of hydrogen, the DCO route is favoured

(NSF, 2011; Pearlson, 2011). When the DCO process is used, feedstock carbon is lost by oxidation and, as

a result, the yield of the process is reduced. When hydrogen inputs are imported in the HDO process,

although the yields are generally higher, the cost and sustainability of the imported hydrogen has to be

assessed. These two alternative routes are simplified in Figure 1-9. The deoxygenation can be carried

out either biologically or thermo-chemically. However, the trade-off between hydrogen inputs and

process yields remains unchanged. In either of these two drop-in biofuel processes, the ultimate

Ethanol

Butanol

Biodiesel

Crude oil

R² = 0.9944

0

5

10

15

20

25

30

35

40

45

0 0.1 0.2 0.3 0.4 0.5 0.6

Ene

rgy

de

nsi

ty L

HV

(M

J/L)

O/C molar ratio

(35% O)

(21.5% O)

(11% O)

(No O) Drop-in biofuel


July 2014 32

objective is not only to deoxygenate biomass intermediates but to also enrich them in hydrogen and

thus elevate their low H/C ratio to the level of finished petroleum transportation fuels (H/C of about 2).

Figure 1-9: Simplified representation of carbohydrate deoxygenation mechanisms

1.7.2 The Hydrogen to Carbon ratio

The hydrogen to carbon ratio (H/C ratio) is used in the petroleum and coal industry to indicate how

hydrogen rich and energy dense are various fossil feedstocks. In the production of drop-in biofuels one

of the main objectives is to elevate the low H/C ratio of the biomass feedstock to that of diesel, jet and

gasoline fuels which have H/C ratios close to 2. During combustion, the oxygen within the biomass

consumes hydrogen and thus reduces its effective H/C ratio. Thus, using a biomass feedstock where the

main elemental components are hydrogen carbon and oxygen, the H/C ratio must account for the

relatively high level of oxygen (in contrast to petroleum feedstocks which contain practically no oxygen)

as each oxygen atom consumes two hydrogen atoms to form a water molecule (H2O) that contributes

no energy to the combustion system (Vennestrøm et al., 2011). Thus, the “effective” H/C ratio for

oxygenated biomass feedstocks, Heff/C, is calculated by Equation 1.

Heff/C =

𝒏(𝑯)−𝟐𝒏(𝑶)

𝒏(𝑪) where n=number of atoms of each element

A Hydrocarbon e.g., Butane C4H10

A Carbohydrate e.g., Glucose C6H12O6

-H2O

-CO2

- O2

1. Insert Hydrogen

2. Oxidize Carbon

“Sacrificing” feedstock carbon

High H/C ≈ 2

H2


July 2014 33

Equation 1: The Effective Hydrogen to Carbon ratio

Highly oxygenated and hydrogen-poor biomass intermediates such as sugar and cellulosic biomass have

low Heff/C ratios. Glucose has an Heff/C ratio of zero meaning that all hydrogen in the substrate is

consumed by its abundant oxygen atoms. When the most common biofuel intermediates as well as

target drop-in diesel molecules are listed along with their Heff/C ratios in a “staircase” depiction, (Figure

1-10) it is apparent that, the wider the gap between Heff/C of a feedstock and a target product molecule,

the more processing and hydrogen input efforts (“steps”) that have to be taken to reach a target drop-in

biofuel H/C situated at the top of the staircase. Thus, for example, a lipid feedstock used by the

oleochemical drop-in biofuel platform will be favoured over a lignocellulosic biomass feedstock for drop-

in biofuel production (Forsberg, 2009).

Although the “staircase” and Heff/C concept (depicted in Figure 1-10) are useful "rules of thumb" to help

assess the suitability of "biomass materials" as feedstocks for drop-in biofuels, these concepts need to

also consider biomass intermediates that are rich in both hydrogen and oxygen. These types of

intermediates include monoalcohols such as ethanol and butanol which, although they have a Heff/C

ratio of 2, are still too oxygenated to be considered as drop-in biofuels. Other intermediates such as

lignin, although less oxygenated than sugars, are still several steps away from the drop-in Heff/C target of

2. Typically, alcohol feedstocks have been used in in less conventional processes such as alcohol-to-jet

fuel processes (discussed further in Chapter 5). In these processes, although the alcohols benefit from a

high Heff/C ratio they still need to be further deoxygenated (e.g. by using more hydrogen inputs) in order

to produce hydrocarbons that are oxygen-free and suitable as drop-in biofuel blendstocks. It should also

be noted that if these alcohol and lignin feedstocks are derived from biomass they will also require some

type of pre-processing before they are available.

As the source, sustainability and cost of hydrogen will likely play an important role in any future drop-in

biofuel sector it is worth discussing how hydrogen is currently produced and used in the oil and gas

based industries.


July 2014 34

Figure 1-10: The effective Hydrogen to Carbon ratio “staircase”

1.8 Hydrogen in the petroleum Industry

It is apparent that the commercialization of drop-in biofuels will be heavily dependent on the availability

and price of hydrogen (H2) inputs in order to elevate biomass H/C ratios. Thus, it is important to better

understand the market trends and potential competition for this key resource. The current major global

use of industrial grade hydrogen is for petroleum refining and ammonia fertilizer production. In the

petroleum industry hydrogen is used to decontaminate (desulfurize and “denitrogenize”) crude oil and

to upgrade (“crack”) heavy oils to make lighter fuel products. Currently, and for the foreseeable future,

the petroleum industry is and will be constrained by hydrogen availability due to the increasing need for

hydrogen to upgrade crude oils of decreasing quality to satisfy market demand for increasingly pure and

light petroleum products.

1.8.1 Declining quality of crude oil

Crude oil is not a homogeneous or consistent feedstock. Petroleum quality varies considerably between

regions and over the lifetime of a reserve. The density (expressed as the “American Petroleum Institute

Sugar 0

0.2 Lignocellulose

0.4

Diesel 2.0

1.8 Lipids

Oleochemical feedstock

Thermochemical feedstock

Biochemical feedstock

‘Drop-in’ biofuel

1.0

0.6

0.8

1.2

1.4

1.6

Lignin


July 2014 35

(API) index”) and sulfur content (expressed as % wt) are the most commonly quoted crude oil quality

parameters and they are viewed as approximate estimations of carbon content and energy content

(heating value). Good quality oil is characterized as “light” (low density, high API) and “sweet” (low

sulfur content) (Bunting et al., 2010).

Recent trends show crude oil feedstock decreasing in quality over time as it averages sourer (high sulfur)

and heavier indices. For example, over the last 15 years, US refineries have been processing increasingly

heavy and sour crude oil. For the oil industry, these trends translate to higher energy use and GHG

emissions per unit of crude oil processed. In the US, the average energy consumption per crude oil

processed has increased by more than 50% from 2001 to 2011 (S&T2, 2013). This trend of decreasing

crude oil quality is expected to continue and, as shown in Figure 1-11, any forecasted growth in global

oil reserves to the year 2020 has been projected to come from sour, heavy or acidic (high total acid

number or TAN) crudes (US EIA, 2006). It is likely that refiners will have to adapt by utilizing more

complex unit operations and using higher energy and hydrogen inputs to process these crudes

Figure 1-11: Purvin & Gertz forecast for world crude oil quality (Source: data from US EIA, 2006)

Oil refineries upgrade heavy crude oil to light products (such as naphtha) by hydrocracking the heavier

distillates, thus increasing the hydrogen input requirements of the refining facility. Cracking of heavy

petroleum fractions can also be performed in catalytic crackers, which require no hydrogen inputs.

0

10

20

30

40

50

60

70

80

90

100

1990 1995 2000 2005 2010 2015 2020

mill

ion

bar

rels

pe

r d

ay

High TAN

Heavy Sour

Light Sour

Light sweet


July 2014 36

However, catalytic crackers convert a significant proportion of the feedstock to tar residues which

translates to lower conversion yields. Thus, this approach is not favoured, especially when crude oil

prices are high (generally above $60/bbl) (US EIA, 2007). This hydrogen vs yield dilemma has also been

discussed in the context of upgrading biomass to drop-in fuels and it will be an interesting, common

challenge for both petroleum and biomass refineries. (Oil refining operations are described in further

detail in Section 1.9).

Oil sands are among the heaviest crudes currently being processed within the global pool of petroleum

reserves. Oil sands are relatively “young” oil deposits that are tightly trapped between geological

sediments. They comprise a mixture of heavy bitumen oil and sand that requires high energy inputs to

process – for extraction, separation and refining – to finished fuels. To give some sense of the scale of

operations for oil sands utilization, Alberta’s production is projected to increase from a current level of

approximately 1.5 million barrels/day of synthetic crude oil (cf. world daily demand is 90 million barrels

(IEA, 2011)) to 3.5 million barrels a day by 2020 (Alberta Government, 2007). From an emissions’

perspective, oil sands extraction and refining result in significantly higher impacts than does the

production of conventional liquid petroleum fuels. According to the US National Energy Technology

Laboratory, Canadian Oil sands WTT (well to tank) GHG emissions amount for about 34 kg CO2

equivalent/ MMBtu LHV (Low Heating Value) diesel, which is more than double that of the benchmark

crude oil WTI (Western Texas Intermediate) (Gerdes & Skone, 2009).

Coal to liquid (CtL) conversion to make transportation fuel is another fossil-based alternative that has

been used during situations where access to petroleum has been restricted. CtL has been categorised as

the production of Synthetic Liquid Hydrocarbons (SLH), indicating that these liquid hydrocarbons have

been synthesized from non-liquid hydrocarbon sources (e.g. coal or natural gas). This technology can

produce functional equivalents to jet, diesel and gasoline and relies on catalytically condensing syngas to

hydrocarbon liquids, a process widely known as Fischer-Tropsch (FT) synthesis. The liquids resulting

from the FT process vary in properties and the process can be optimised to produce products fitting

gasoline jet or diesel fuel specifications. The chemistry of these liquids differs from traditional petroleum

fuels in that they contain fewer aromatics and therefore are generally better suited for blending with

heavier distillates (diesel and jet) than gasoline.

The company South Africa Synthetic Oil Ltd. (Sasol) is the world’s largest CtL manufacturer with

extensive experience stretching back to the 1950s as a consequence of South Africa having abundant


July 2014 37

coal and limited petroleum resources (Bauen et al., 2009). In their Ctl process, coal is gasified and the

resulting gas (syngas) catalytically condensed to liquid fuels. Sasol’s coal-to-liquid plant in Secunda,

South Africa has increased production up to its current processing capacity of 40 million metric tonnes

of coal per year being used to produce approximately 150,000 barrel oil equivalents (on an energy

basis) of liquid hydrocarbon fuels per day. In 2009, after a 7 year certification process, ASTM approved

Sasol’s semisynthetic jet fuel blends (containing 50% coal-derived and 50% petroleum-derived) (Bauen

et al., 2009).

In a related area, the US Navy and Air Force both have prioritized the procurement of Synthetic Liquid

Hydrocarbons (SLH) as a means to reduce reliance on petroleum imports (SAFN, 2012). In 2006, the US

Air Force initiated testing of FT-jet fuel blends in all aircraft types. However, the Energy Independence

Security Act of 2007 banned the procurement of fuels that are more carbon intensive (higher GHG

emissions) than existing petroleum fuels. This Act essentially precluded coal-derived FT liquids from US

vehicles due to their high GHG emissions (Bauen et al., 2009). Unless options such as Carbon Capture

and Storage (CCS) can be employed, low-emissions biomass-derived SLH will likely be only alternative

fuel choice using this type of conversion technology.

It is likely that unconventional sources of refinery feedstock such as heavy oil, oil sands and coal

liquefaction will increase the environmental and economic costs of producing current transportation

fuels. These “less than ideal” fossil feedstocks are generally hydrogen deficient and, just like biomass

feedstocks, have much lower hydrogen to carbon ratios than “high quality crudes” such as Saudi Arabian

light and sweet. Similarly to the previously described “staircase” arrangement, fossil feedstocks can also

be compared for their relative processing and hydrogen requirements with an H/C staircase (Figure

1-12). Coal, which has an H/C ratio of about 0.5 requires 3 times more hydrogen inputs than do lighter

crudes which have an H/C ratio of about 1.5 (Forsberg, 2005). These numbers indicate that there is a

near linear relationship between H/C ratio and hydrogen requirements regardless of the type of

feedstocks to be converted to finished fuels. This linear relation is a recognised rough-rule-of-thumb

used by the petroleum refining industry but it has to be adapted to account for the presence of

heteroatoms such as sulfur and nitrogen, which, just like oxygen, consume hydrogen during upgrading

and refining (Forsberg, 2005).


July 2014 38

Figure 1-12: The hydrogen to carbon “staircase” for fossil fuel feedstocks

1.8.2 Demand for low sulfur and light petroleum products

As petroleum reserves become heavier and sourer, an opposing trend is occurring in petroleum markets

where increasingly light and low-sulfur products are in demand. Globally, light products and particularly

middle distillates (the main blendstock for jet fuel and diesel) are in increasingly higher demand than

heavier fractions such as fuel oil and bunker fuels. The increase in demand for prime quality

transportation fuels is mainly a result of the increased demand for long distance transportation fuels (in

non-OECD countries in particular) and the tightening sulfur emission regulations in road and marine

transport (in OECD countries especially).

While electric and natural gas vehicles are proposed as alternatives to petroleum-based light duty

vehicles, these alternatives are not viable for longer distance transportation modes such as air travel

and shipping. These latter, more inelastic and oil-dependent sectors are also where most of future

petroleum fuel demand growth is expected to occur. Another long distance transportation mode that is

projected to grow is truck/lorry freight transport, particularly in emerging non-OECD (Organization for

Economic Cooperation and Development) economies such as China and India. In these countries

0.4

Diesel 2.0

1.8 Light crude

1.0

0.6

0.8

1.2

1.4

1.6

Coal

Oil sands

Heavy crude


July 2014 39

domestic trade is now growing faster than export trade and goods are increasingly transported by road

rather than by air or sea (IEA, 2012).

The petroleum feedstock used to make diesel fuels as well as higher grade heating oil and bunker fuels is

the “middle distillates” petroleum fractions. Demand for middle distillates, more commonly called

gasoil, is expected to grow exponentially over the next few decades as current fuel substitution trends

increase, including fuel switching from gasoline to diesel, from heating oil to natural gas/electricity and

from bunker fuels to higher quality marine gasoil. All of these trends, sometimes collectively referred to

as the “dieselification” trend, favour increased diesel and gasoil production. According to the IEA (2011),

gasoil alone accounts for almost 40% of total forecast growth in oil demand through to 2016 while its

share of total petroleum product demand will climb steadily to 30% by the same year. The overall

tendency for markets to grow mostly around the middle part of the barrel (middle distillates) is

illustrated in Figure 1-13, where the demand growth for middle distillates is clearly evident. In

aggregate, this growth represents 46% of the total demand growth through to 2017. During the same

time period, heavy fuel oil demand is expected to show no or negative growth.

Figure 1-13: World oil demand by product, 2010-2017


July 2014 40

(Source: IEA, 2012b) Tightening fuel emission regulations are the second factor that will drive increased demand for light and

low sulfur petroleum fractions as opposed to heavier and more sour fractions. Sulfur emission limits

continue to be reduced, particularly for road and marine transport applications (IEA, 2011a; IMO, 2012).

Bunker fuels, once thought of as the refining sector’s “sink hole” for the lowest quality refinery fractions,

are now also tightly regulated for their sulfur emissions by the UN’s International Maritime Organization

(IMO), as described in the International Convention for the Prevention of Pollution from Ships

(MARPOL). MARPOL’s Annex VI, in force since 2010, has focused on the establishment of “Sulfur

Emission Control Areas” (SECAs) in high traffic European jurisdictions such as the Baltic and North Seas.

Other SECAs have been designated in North America (in effect since August 2012) and even more are

expected to be introduced in high shipping traffic hot spots in South American and the Pacific Rim (IEA,

2011a; IMO, 2012). In practice this means that, when operating within a SECA, a ship must only burn low

sulfur fuels. Annex VI has been ratified by 63 countries, which account for some 90% of the gross

tonnage of the world’s merchant fleet. The fuel sulfur content limits under MARPOL are currently 1 wt%

in SECAs and 3.5% outside SECAs. These limits are to be further decreased over the next decade to 0.1%

in SECAs by 2015 and to 0.5% worldwide by 2020-2025, to the extent of sufficient availability of

compliant low sulphur fuels (IEA, 2011a; IMO, 2012).

Sulfur emission control is also being applied to road transport as indicated by the fuel specification

trends in national fuel standards, particularly in the EU and the US. In the EU, the specifications for

maximum allowable sulfur in diesel (EN 590) dropped from 2000 ppm in 1994 to 10 ppm in 2009. Similar

trends have been observed in the US where, as shown in Figure 1-14, maximum allowable sulfur

contents of fuels have dropped over the last decade from hundreds to thousands of ppm down to only a

few ppm for all types of gasoline and diesel fuels (US EIA, 2006).

Interestingly, the aviation sector has always been exempt from fuel emission regulations and the sulfur

limits in aviation fuels can be as high as 3000 ppm although, in practice, jet fuel sulfur levels currently

average only about 600 ppm worldwide (King, 2012). On the other hand, a recent analysis concluded

that although desulfurizing jet fuels would reduce health impacts, it would increase these fuels’ climate

impact (because of removing cooling sulfate particles) and thus the costs and benefits came out to be

broadly even (Gilmore et al., 2011). It is also likely that low maximum sulfur specifications for all other

fuels will at least indirectly affect the sulfur content of jet fuels. For example high-sulfur jet fuel suppliers


July 2014 41

may have to find alternative distribution systems if pipeline operators, concerned about sulfur

contamination to diesel and gasoline, stop accepting high-sulfur fuels (US EIA, 2006).

Figure 1-14: Sulfur Content Specifications for U.S. Petroleum Products, 1990-2014 (Source: US EIA, 2006) In summary, the combination of declining crude oil quality and increasing demand for lighter and more

refined petroleum based fuels will create a need for increased, global crude oil upgrading capacity. As a

result and as described in the next section, large amounts of hydrogen are going to be required.

1.9 Oil refining basics (emphasis on hydrotreating and hydrocracking)

As oil refineries will increasingly have to adapt to hydrogen consuming processes, the following section

provides a brief overview of oil refining basics with emphasis on hydrotreatment and hydrocracking and

their role in removing sulfur from sour crude oils and in “cracking” heavy crudes to lighter products.

1.9.1 Crude oil

Crude oil is organic material which has been converted to a carbon-rich liquid over millions of years,

under conditions of high temperature/pressure in between geological sediments. It comprises a mixture

of hundreds of hydrocarbon molecules ranging in size (1 to 300 carbon atoms) and structure. In terms of

structure, the hydrocarbon molecules that constitute crude oil are generally classified as:

Paraffins (e.g., n-octane, n-cetane, CnH2n+2)

0

500

1000

1500

2000

2500

1990 1995 2000 2005 2010 2014

Sulf

ur

pp

m Reformulated Gasoline

Traditional Gasoline

Diesel

Non-road diesel


July 2014 42

Isoparaffins (or branched alkanes, e.g., isocetane, CnH2n+2)

Olefins (e.g., ethene, CnH2n)

Cyclic paraffins (cycloalkanes or naphthenes are alkanes containing one or more saturated

carbon atom rings)

Aromatics (hydrocarbons containing one or more benzene rings)

Impurities such as sulfur, nitrogen and metals are also present in crude oil.

1.9.2 Oil refining

Oil refining refers to a complex system of industrial processes which converts crude oil to fuel grade

liquids and value-added products such as chemicals and polymers/plastics. The simplest refinery process

can be thought of as one where crude oil is heated and the different fractions collected and cooled to

form (from lightest to heaviest) liquefied petroleum gas, gasoline, naphtha, kerosene, gas oil and fuel

oil. The conversion entails a number of processes which fall under the three main categories of

distillation, upgrading and blending. As simplified in Figure 1-15, and starting from the left side of the

figure, a “generic” oil refinery consists of a distillation unit, which fractionates crude oil into product

streams (also known as “cuts”) according to their boiling point ranges. The lighter cuts (lower boiling

point fractions) such as naphtha and “straight run gasoline” are typically used for gasoline grade fuels

while the extremely light, mostly gaseous cuts (shown as “light ends” in Figure 1-15) are used for more

value-added products. For example, light olefins (e.g., ethylene) and light aromatics (e.g., xylene) which

are recovered near the top of the distillation column, are precursors to a number of high value polymer

and plastics such as polyethylene and polyurethane. The middle streams from the distillation tower

(after desulfurization) such as kerosene and gas oil are used for jet and diesel fuels respectively. These

middle streams are also collectively known as “middle distillates”. The heavier, higher boiling point

streams go to residual fuel (which forms part of marine and heating fuel blends) or are sent to cracking

facilities such as hydrocracking and Fluid Catalytic Cracking (FCC) units, to be upgraded to lighter, higher

value products. The vacuum distillation tower is able to process extremely heavy residue into lighter,

FCC feed and to heavier, coker feed. Cokers yield lighter FCC feed and very heavy and dense coke

suitable as combustion fuel and typically used for power generation (US EIA, 2007).


July 2014 43

Figure 1-15: Simplified diagram of an oil refinery (Source: adapted from US EIA, 2007)

Downstream from the distillation tower are all the upgrading units which perform the three main

functions of: a) breaking apart heavy, low-valued molecules into lighter more valuable streams; b)

rearranging molecules to improve performance or emission specifications (e.g. reforming gasoline cuts

to boost their octane rating); and c) removing undesirable materials such as sulfur and other impurities.

The upgrading units represent the biggest capital expense in a refinery. They are also costly to run as

they operate at high pressures and often use catalysts and large amounts of hydrogen inputs (Figure

1-15). Hydrocrackers, hydrotreaters and FCCs are key unit operations for upgrading the increasingly

marginal oil feeds. They are also important units to consider in the current study because they dictate

the hydrogen demand of the refinery and are a possible “insertion point” for upgrading of drop-in

biofuel intermediates to deoxygenated hydrocarbon fuel blendstocks. Blendstock is a term used to

describe a fuel stock that is transported to blending terminals where they are mixed to produce finished

fuels meeting specific market specifications.

1.9.3 Fluid Catalytic Cracking (FCC)

The fluid catalytic cracking (FCC) unit is predominantly a gasoline production unit and typically supplies

roughly 50% of the gasoline produced in a standard refinery. FCC units crack heavy molecules to mostly

Crude oil

Distillatio

n

to

we

r

Vacuum unit

Light ends

Heavy ends

Reformer Gasoline

Jet

Hydrotreatment Diesel, Jet

Hydrocracker Diesel, Jet

Gasoline

Fluid catalytic cracking

Coker Coke

Hydrotreatment Gasoline

Diesel, Jet

Pro

du

ct blen

din

g

DISTILLATION (CATALYTIC) UPGRADING BLENDING

Naphtha

Kerosene, Gas oil

Fuel oil

Gasoline


July 2014 44

gasoline and some middle and heavy distillates. They do not use hydrogen and, as a result, higher

amounts of coke are formed on the surface of the catalysts. Therefore, an operating FCC always has a

regenerator attached to it (Figure 1-16) where the coke on the spent catalyst is burned off and the

regenerated catalyst is re-injected into the main reactor. Since biomass substrates also deposit a lot of

coke on catalysts, the FCC configuration outlined in Figure 1-16 is also popular in pyrolysis-type

thermochemical conversions of biomass to drop-in biofuels.

Due to the global reduction in demand for gasoline and increasing demand for diesel, the need for more

FCC units, which are predominantly gasoline producing units, is declining. Even in the US, the world’s

biggest consumer of gasoline (both in absolute volumes and as a share of the country’s total

transportation fuel mix), FCC use is declining and about 20% of the FCC units in US refineries are

currently idle (US EIA, 2013). It should also be noted that their decommissioning is expensive. Thus, this

situation may represent an opportunity for thermochemically based drop-in biofuel processes, such as

pyrolysis, to better utilize idle FCC units. However, as discussed in more detail in chapter 3, this comes

with several logistic and processing challenges.

Figure 1-16: Simplified representation of a fluidized catalytic cracker (Adapted from Corma et al., 2007)

Spent Gases

Regenerator

Reactor

Rise

r Spent

Catalyst

Fee

d

Regenerated

Catalyst Steam

Air


July 2014 45

1.9.4 Hydrotreating

Hydrotreating is a hydrogen consuming process that removes sulfur and other impurities from

petroleum product streams. The process involves high temperatures and pressures as well as specialized

catalysts. During hydrotreatment, the hydrogen reacts with the sulfur to form hydrogen sulfide gas

which is then sent to the sour gas treatment unit of the refinery. The reactors are mostly fixed bed units

and the catalysts are usually cobalt or molybdenum oxides or alumina. However, they can also contain

nickel and tungsten (US EPA, 1995). Catalysts are replaced or regenerated at an offsite facility after

months or years of operation. Hydrotreating facilities range in size between 50,000-150,000 bpd (barrels

per day) and, depending on the quality of the feed, their hydrogen consumption can range between 7

and 285 Nm3/barrel (bbl), pressures between 0.7 and 15.5 MPa and temperatures between 300 and 450

°C. Heavy and high sulfur feeds require the use of the top range of these conditions to be fully

desulfurized (J. Speight & Özüm, 2002). Hydrotreating is also used to remove nitrogen impurities in a

process known as HDN (hydrodenitrogenisation). However, HDN is typically a minor side reaction within

HDS as nitrogen impurities are typically only present at low levels in petroleum liquids.

1.9.5 Hydrocracking

As the name implies, hydrocracking is a hydrogen-adding process which breaks apart low value, heavy

petroleum molecules to high value light molecules. A simplified depiction of the process is shown in

Figure 1-17. Hydrocracking can be viewed as a more severe form of hydrotreating. In a refinery the main

function of this unit operation is to process the low quality heavy distillates obtained from the

distillation tower, the FCC and coker units, and to convert them to desulfurized finished fuels (mostly for

diesel and jet applications). In the emerging world of increasing fractions of heavier and more sour

crude oils where low sulfur fuels are mandated and demand growth is predominantly in the diesel and

jet fuel markets, hydrocracking will play an increasingly pivotal role as modern oil refineries adapt

operations to meet emerging market fuel trends.


July 2014 46

Figure 1-17: Simplified depiction of a hydrocracking reaction (adapted from Sotelo-Boyas, Trejo-Zarraga, & Hernandez-Loyo, 2012)

Hydrocrackers operate under very high pressures (1200 to 2000 psig) and most catalysts consist of a

crystalline mixture of silica-alumina with small amounts of noble metals such as platinum and palladium.

These catalysts are typically regenerated offsite every 2-4 years and they are sensitive to water and

heteroatom impurities such as sulfur and nitrogen. Water is usually removed by passing the feed stream

through silica gels or molecular sieves prior to feeding to the hydrocracker. Sulfur and other impurities

are often removed by hydrotreating the feed stream prior to feeding (US EPA, 1995). An average

hydrocracker has the capacity to process about 60,000 barrels of feed per day. It consumes about 57

Nm3/bbl of hydrogen and costs over USD 400 million to build (Gary et al., 2007).

1.10 Hydrogen demand in the oil industry

As has been documented, hydrogen will be increasingly required to convert heavy and sour crude oil to

high value, low sulfur finished fuels. Thus, hydrogen demand for oil processing will continue to increase

and the sector will continue to be the world’s biggest user of industrial grade hydrogen (Levin &

Chahine, 2010; Mohamed et al., 2011). Hydrotreating capacity in the US alone is expected to double by

2030 to 27 million barrels a day, from 14 million barrels a day in 2004 (Figure 1-18). This doubling of

CH3

CH3

CH3

CH3 CH3

CH3

CH3 CH3

CH3

CH3

CH3

CH3

CH3

15 MPa 400 °C

NiMo/γ-Al2O3

Heavy crude molecule

Gasoline range molecule Diesel range molecule


July 2014 47

hydrotreating capacity will be associated with an equal increase in hydrogen demand to supply US

petroleum refineries.

Figure 1-18: Hydrotreating capacity in the US (Source: US EIA, 2006)

1.11 Hydrogen generation in oil refineries

Currently, most refineries produce hydrogen onsite using steam reforming of methane (SMR). The two

main reactions involved in this conversion are:

CH4 + H2O → CO + 3H2

CO + H2O → CO2 + H2

In the first reaction, methane reacts with high temperature steam to form syngas (CO and H2). In the

second reaction (also called the “water-gas shift” reaction) the CO produced from the first reaction is

converted to hydrogen. The methane is usually derived from natural gas, although sometimes the

refinery off-gases can also be used as a source of methane and CO to feed the hydrogen-producing

0

5

10

15

20

25

30

1990 1995 2000 2004 2010 2015 2020 2025 2030

mill

ion

bar

rels

pe

r d

ay


July 2014 48

reactions. A typical refinery steam reformer can process about 65,000 m3/h of natural gas and costs

around USD 200 million to build (Muellerlanger et al., 2007).

As mentioned earlier, refineries that have limited access to natural gas may gasify some of the residual

crude oil cuts to make hydrogen. However, this is a highly energy intensive process and the use of a

petroleum feedstock to make hydrogen as opposed to natural gas is unlikely given the current trends of

increasing oil prices and dropping natural gas prices. The price of natural gas has been declining for the

last few years (Figure 1-19) and with the ongoing exploration and use of new technologies to help

extract “unconventional” natural gas (e.g. shale gas) this low price is expected to continue for some

time. Over the same time, as shown in the graph (Figure 1-19), the price of crude oil has increased, thus

further increasing the cost differential between the two fuel commodities.

Figure 1-19: Crude/Natural gas price differentials Source: Data from (IndexMundi, 2013)

0

20

40

60

80

100

120

140

160

0

100

200

300

400

500

600

Jan

-91

Feb

-92

Mar

-93

Ap

r-9

4

May

-95

Jun

-96

Jul-

97

Au

g-9

8

Sep

-99

Oct

-00

No

v-0

1

De

c-0

2

Jan

-04

Feb

-05

Mar

-06

Ap

r-0

7

May

-08

Jun

-09

Jul-

10

Au

g-1

1

Sep

-12

USD

/ b

arre

l

USD

/ 1

00

0 m

3

Natural Gas Price (USDollars per thousand cubicmeters of gas)

Crude Oil (petroleum);Dated Brent Price (USDollars per Barrel)


July 2014 49

1.12 Hydrogen use in the ammonia fertilizer industry

While oil refining utilises the largest amounts of hydrogen, the ammonia fertilizer industry is the second

largest global consumer of hydrogen. Demand for fertilizer continues to increase and prices for most

agricultural commodities are currently at record highs (the FAO food index has increased from 100 units

in 2003 to 250 units in 2013). Diets in countries such as China and India are becoming more

“westernized”, resulting in an increased demand for fertilizer. Ammonia, the most widely produced

fertilizer commodity in the world, is produced by chemically reducing atmospheric molecular nitrogen

(N2) to ammonia gas by reacting it with hydrogen. The hydrogen (gas) typically comes from steam

reforming natural gas. Thus, natural gas is one of the most critical feedstocks for the ammonia industry.

The global capacity for ammonia manufacturing was 161.3 million tonnes (on a nitrogen basis) in 2011

and it is expected to reach 182.2 by 2016 (Food and Agriculture Organization (FAO), 2012). Most of the

capacity growth is expected to be in Asian countries, mainly in Southeast Asia while significant growth is

also expected in Eastern Europe.

1.13 Biomass as a source of hydrogen

Although more than 90% of the world’s industrial hydrogen is produced by steam methane reforming

(SMR) from fossil natural gas, there are numerous technologies for producing hydrogen from renewable

resources. For biomass conversion processes, the most available resource from which to derive the

renewable hydrogen is the biomass itself. The most effective way of making hydrogen from biomass is

to use gasification followed by steam reforming of the resulting syngas. However, this process is less

than half as efficient (on an energy basis) as the conversion of natural gas to hydrogen, which

approaches efficiencies of about 90%. Holladay et al. (2009) reviewed the efficiencies of technologies

related to hydrogen production from both fossil and renewable biomass resources. They concluded that

the efficiency of natural gas steam methane reforming is excellent. The same report described how all

fossil-based technologies will continue to be more efficient than renewable hydrogen technologies for

the foreseeable future. In related work, when Lau et al. (2003) performed a techno-economic analysis to

compare the cost of hydrogen derived from the gasification of either bagasse, switchgrass or palm

nutshell with that of natural gas (Figure 1-20) it was only competitive at a very large scale.


July 2014 50

Figure 1-20: Estimated Cost Hydrogen Production from Biomass and Natural Gas Feedstocks (Source: Lau et al., 2003)

From these estimates, it can be concluded that, at the current natural gas prices of about $4/GJ, a 1000

t/day bagasse-to-hydrogen facility would generate hydrogen that would cost almost twice as much as

that derived from natural gas. This cost disparity is even more problematic when the average cost of

biomass, which was $30/dry tonne (dt) for this study, is considered. Most recent techno-economic

analyses estimate much higher biomass costs, in the range of $60-75/dt or higher. Most studies

acknowledge that using biomass to produce hydrogen is a not an ideal option in terms of energy

efficiency and associated GHG savings (Levin & Chahine, 2010).

It is apparent that a key challenge for developing drop-in biofuels will be finding cheap, sustainable

sources of hydrogen. However, there is an opportunity to develop enhanced compatibility and

leveraging opportunities with the current oil refinery infrastructure. It is also clear that long distance

transportation such as aviation, shipping and trucking has few alternative options other than liquid

biofuels. A key consideration will be the efficiency of the conversion processes used to deoxygenate

biomass feedstocks. The trade-offs that will be encountered and the production and process challenges

for each of the drop-in biofuel options are reviewed in the next section of the report.


July 2014 51

CHAPTER 2: THE OLEOCHEMICAL PLATFORM

Most “drop-in” biofuel demonstrations to date have used material derived from oleochemicals. The

oleochemical platform is based on lipid feedstocks such as vegetable oils or other bio-derived fats such

as tallow and algal oils. These lipid based feedstocks have been the pioneers in the manufacture and

demonstration of drop-in biofuels primarily because they contain low amounts of oxygen and have a

high hydrogen-to-carbon (Heff/C) ratio – they are already close to drop-in fuels – compared to sugar or

cellulosic feedstocks. Conventional lipid based biofuel known as “biodiesel” is produced by esterification

of triacyl glycerides (TAGs) to produce fatty acid methyl esters (FAME). However, biodiesel is not fully

compatible with existing petroleum infrastructure. A further hydroprocessing step is required to convert

lipids into deoxygenated hydrocarbon drop-in biofuels typically known as hydrotreated esters and fatty

acids (HEFA). This hydrogen-requiring process represents the only route to date that has been used to

deliver commercially meaningful amounts of drop-in biofuels. As noted earlier, HEFA biofuels have been

the main aviation biofuel used for test flights carried out by the US Navy and many commercial airlines.

2.2 Process overview

2.2.1 Conventional oleochemical-based biofuel (esterified fatty acids)

As shown in Figure 2-1, the esterification of vegetable oils or other bio-derived fats to make biodiesel

involves the reaction of a TAG lipid with methanol in presence of a base, such as NaOH, or acid, such as

H2SO4, to form FAME and produce glycerol as a by-product. This conversion is relatively simple and,

unlike HEFAs, there is no requirement for specialized catalysts or hydrogen (H2) or for high pressures

and temperatures. Rather, production of FAME can be performed at various scales ranging from small

“backyard-type” units all the way to large scale biodiesel-manufacturing facilities such as the 100 million

gallon per year (380 MLPY) Imperium Renewables biodiesel facility in Washington State, USA.

http://www.imperiumrenewables.com/index.html


July 2014 52

Figure 2-1: Triglyceride to Biodiesel (FAME) reaction

The major disadvantage of FAMEs when compared to HEFAs is that they are not a “drop-in” biofuel. As

is shown in Figure 2-1, there is still an appreciable amount of oxygen present in the final FAME product

and this imparts polar and hydrophilic chemistry that inhibits full compatibility with existing fuel

infrastructure: a) it contains a lower energy content than oxygen-free hydrocarbon fuels; b) has a higher

cloud point temperature which limits the applicability of the fuel in cold climates; c) reacts with water

and can contaminate petroleum blends; d) reacts with metal surfaces and sticks to them and/or causes

corrosion; and e) reacts with itself which reduces fuel storage life. When the properties of petroleum

diesel and biodiesel are compared (Table 2-1) most of the biodiesel deficiencies are directly or indirectly

related to its oxygen content, although the lower sulphur content of biodiesel is an exception and is one

of the few advantages biodiesel has over petroleum-derived diesel. Aromatics are missing from all

oleochemically derived biofuels (both HEFA and FAME) except tall oil.


July 2014 53

Table 2-1: Properties of HEFA petroleum diesel and FAME biodiesel.

Properties HEFA

Renewable Diesel

Fossil Diesel EN 590

(summer

grade)

FAME Biodiesel

(from rape

seed oil)

Density at 15 °C (kg/m3) 775 - 785 835 885

Viscosity at 40 °C (mm2/s) 2.5 - 3.5 3.5 4.5

Cetane number 80 - 99 53 51

Distillation range (°C) 180 - 320 180 - 360 350 - 370

Cloud point (°C) −5 to −25 −5 −5

Heating value, lower (MJ/kg) 44.0 42.7 37.5

Heating value, lower (MJ/L) 34.4 35.7 33.2

Total aromatics (wt-%) 0 30 0

Polyaromatics (wt-%) (1) 0 4 0

Oxygen content (wt-%) 0 0 11

Sulfur content (mg/kg) < 10 < 10 < 10

Lubricity HFRR at 60 °C (mm) < 460(2) < 460(2) < 460

Storage stability Good Good Challenging

(1) European definition including di- and tri+ -aromatics (2) With lubricity additive Source: (Aatola et al., 2008)

Another critical yet often overlooked challenge of biodiesel is its limited compatibility with petroleum in

pipelines. Biodiesel transported through conventional petroleum pipelines can mix with the “water

plugs” which are inserted into the pipeline to separate different petroleum liquids from each other

when they are transported through the same pipeline at different times. Biodiesel can also stick to

pipeline walls and contaminate jet fuel plugs that follow. As current jet fuel specifications are very

stringent, such a contamination risk prohibits the transport of jet fuel and biodiesel in the same

petroleum pipelines. Biodiesel is typically transported via road/trucks, a practice that adds to the cost

and carbon footprint of the fuel. However, Kinder Morgan Energy Partners (KMEP), one of the largest

pipeline companies in North America, has successfully shipped biodiesel through its Plantation pipeline

network located in the southeastern United States. In this case, jet fuel contamination was prevented

due to the existence of a parallel pipeline that allowed jet fuel and FAME biodiesel to remain


July 2014 54

segregated. The results of this trial were positive and, as of early 2010, KMEP allows the shipment of B2,

B5, and/or B100 in over 8000 miles of pipeline (Bunting et al., 2010).

Due to these compatibility concerns, and several other issues, biodiesel is seldom used neat or as a

finished fuel (100% biodiesel or B100). To try to make use of the current infrastructure and motor

engine compatibility, it is typically blended with conventional diesel at ratios of 5 or 7% volume (B5 or

B7). Currently biodiesel is primarily used in road transportation while its use in jet fuel blends is strictly

prohibited.

2.2.2 Advanced oleochemical (hydrotreated lipids)

Hydroprocessed Esters and Fatty Acids (HEFA) is the term used to describe drop-in biofuels that are

produced by hydrotreating lipids derived from vegetable, algae and animal fats. To distinguish HEFA

from “biodiesel”, the terms “green diesel” or “renewable diesel” are often used. Alternative acronyms

for HEFA renewable fuels include HRV (Hydrotreated Renewable Vegetable oils), HVO (Hydrotreated

Vegetable Oils) and HRO (Hydrotreated Renewable Oils). Other common acronyms used to describe the

type of HEFA used for jet fuels only are HRJ (Hydrotreated Renewable Jet fuel) and bio-SPK (bio-based

synthetic paraffinic kerosene).

Compared to other potential biofuel feedstocks such as sugars and cellulosic biomass, fats are the

simplest to convert to drop-in biofuels because, as discussed earlier, they have low oxygen content and

their chemistry is closer to a hydrocarbon than saccharides or lignins (i.e., their effective hydrogen to

carbon ratio is closer to 2). Despite these benefits, the conversion of fats to HEFA entails significant

capital costs as well as hydrogen inputs compared to the production of biodiesel (FAME). For example,

the capital expenditure for a 2000 tpd HEFA facility, as modeled by Pearlson (2011), is about USD

$2.6/gal ($0.7/L) of installed capacity and the hydrogen use is 3-4% by mass of feedstock compared to

USD $0.8/gal ($0.2/L) and no hydrogen inputs for FAME as modeled by Marchetti et al. (2008) for the

same size facility.

Process chemistry

In a standalone facility, HEFA’s are typically produced in two stages (Figure 2-2) (Pearlson, 2011). During

the first stage the fats are deoxygenated and their double bonds are saturated to create alkanes. The


July 2014 55

second stage involves alkane isomerisation and cracking, bringing the biofuel to a quality specification

that equals or surpasses specifications for conventional petroleum fuels or fuel blendstocks.

Figure 2-2: Simplified Hydroprocessed Esters and Fatty Acids (HEFA) process depicting the 2 stages of hydroprocessing

As depicted in Figure 2-3, during the first stage of HEFA production, a number of chemical reactions take

place with some hydrogen initially used to saturate all of the carbon-carbon double bonds present in the

triacyl glyceride (TAG). More hydrogen is added in the second reaction which removes the propane

backbone of the TAG leaving 3 free fatty acids per TAG molecule. Finally the fatty acids are

deoxygenated either with the addition of more hydrogen (hydrodeoxygenation, where oxygen leaves as

H2O) or with the loss of carbon (decarboxylation, where oxygen leaves as CO2) resulting in the formation

of alkyl chains. During hydrodeoxygenation (HDO) the alkyl chain length is typically preserved whereas

during decarboxylation (DCO) alkyl chains are shortened due to the loss of carbon atoms as CO2. Usually

a combination of the two deoxygenation strategies is used in commercial hydrotreating facilities

(Pearlson, 2011). The ratio of each deoxygenation pathway (e.g., HDO/DCO = 35/65) is of importance to

the hydrotreating operations as it determines the hydrogen consumption, product yields, catalyst

inhibition, gas consumption and heat balance (Egeberg et al., 2010). The tuning of the deoxygenation

Fatty acid feedstock

H

ydro

crac

kin

g/i

som

eri

zati

on

Gasoline

Diesel

Jet

Lights

H

ydro

tre

atm

ent


July 2014 56

pathway ratio can be achieved via catalyst adjustment, depending on the strategic manufacturing

priorities as well as the feedstock and hydrogen costs and the value of the fuel product or blendstock

being produced. For example, the UOP-Honeywell facility opts for more decarboxylation in order to

reduce capital costs while Syntroleum prioritises the preservation of longer carbon chains (higher

product quality) and thus uses more hydrodeoxygenation (Pearlson, 2011).

Figure 2-3: Triacyl glyceride (TAG) deoxygenation process

Source: (Pearlson, 2011)

After the first processing stage, the TAG feedstock has been converted to an oxygen-free, saturated

liquid alkane intermediate. This hydrocarbon liquid can be directly blended in small quantities with

petroleum diesel (Pearlson, 2011). However, it does not meet the specifications of a finished fuel due to

its poor cold flow properties (propensity to freezing at lower temperatures, i.e., relatively high freeze

point and high cloud point). During the second and final process stage, the unbranched long chain

alkanes are cracked and isomerised in a process referred to as “dewaxing” (Pearlson, 2011). Dewaxing

reduces the length and increases branching of the carbon chains, thereby reducing the freezing point of

the resulting finished fuel (Conventional oil refining and the chemistry of cracking and isomerisation are

reviewed in Chapter 1). The mass yield of HEFA liquids from lipid raw material is typically around 80%

but varies according to the feedstock and processing conditions used (Pearlson, 2011; Sotelo-Boyas et

al., 2012). The remaining 20% of material is generally composed of light gases such as propane, methane

and oxygenated gases such as CO2 and CO. Other than CO2, these gases are usually combusted to

provide power for the process. Typical HEFA liquids comprise 3 different fractions corresponding to jet,

diesel and gasoline (or naphtha) blendstocks (Figure 2-2). The distribution of these three liquid product

fractions can be controlled by changing the reaction conditions and catalysts. However, diesel generally

predominates with only a small portion of the liquids in the jet range (Bezergianni et al., 2009; Pearlson,

2011). For example, it has been reported that in a UOP-like (decarboxylation-based) HEFA process, ca.

Triglyceride Hydrogenation

Propane Loss

Deoxygenation

CH2-O-CO-C17H33

CH-O-CO-C17H33

CH2-O-CO-C17H33

CH2-O-CO-C

17H

35

CH-O-CO-C17

H35

CH2-O-CO-C

17H

35

3C17

H35

COOH + C3H8 +3H2 +3H2

+9H2 3C

18H

38 + 6H2O

3C17

H36

+ 3CO2

Decarboxylation


July 2014 57

65 wt% of the incoming vegetable oil gets converted to diesel-range molecules and only ca. 13 wt% to

jet-range; increasing this the jet yield to 50 wt% requires 30% more hydrogen and reduces the overall

liquid fuel yield of the process from 80 wt% to about 70 wt% (Pearlson, 2011). The extra processing

required to maximize jet fuel production imposes extra economic and logistic challenges and, contrary

to common perception, jet fuel does not always command a higher price than diesel. In fact, at the time

of writing jet fuel prices are around USD $3 per gallon while diesel prices are at around USD $3.20 per

gallon ($0.85/L) (IndexMundi, 2013). In the absence of a price premium for jet fuels compared to diesel

fuels, jet fuel would be sold as diesel since jet fuel can be fed to diesel engines (a standard practice of

the US Navy to simplify logistics) but not the other way around. It would be difficult to justify the extra

cost of maximizing HEFA jet yields and the cost of separating jet from diesel fractions unless there is

significant price premium for HEFA jet fuel compared to HEFA diesel. As is discussed in section 2.4.2, it

has been estimated that this premium would need to be USD 2-3 $ /gal ($0.53 – $0.79/L).

2.3 Potential for integration with oil refineries

It has been suggested that hydroprocessed Esters and Fatty Acids (HEFA) platforms can achieve capital

savings by leveraging petroleum refineries by co-processing fats at the same time as petroleum

intermediates in existing hydroprocessing facilities (ConocoPhillips, 2010; Egeberg et al., 2010). This co-

processing strategy is intended to take advantage of existing petroleum refining infrastructure and fuel

off-take networks while also utilizing lower cost hydrogen typically available in oil refineries. As

described earlier, in a standalone facility HEFA is usually produced in two separate stages. In an oil

refinery, this process could be performed in a single combined hydroprocessing stage.

2.3.1 Challenges of hydroprocessing renewable oils in conventional refineries

Hydroprocessing units (hydrocrackers and hydrotreaters) are central components of a typical oil refinery

and the overall economics of operating a refinery is influenced by the performance of these units. As

described below, introducing oxygen-containing renewable oils to hydroprocessing units presents a

number of challenges that must be carefully addressed to ensure continued smooth refinery operation

and profitability.

Renewable oils such as vegetable oils and animal fats are naturally unstable and corrosive due to their

oxygen content and, consequently, problems can be encountered transporting these oils through metal

pipelines within a refinery. Vegetable oils and other natural oils, especially those with a high free fatty


July 2014 58

acid content such as tall oil, can cause severe corrosion of pipes and other metal equipment upstream of

the hydrocracking reactor (Egeberg et al., 2010). These oils must be handled in a similar way to highly

acidic (high-TAN) fossil crudes.

The reactions involved in hydrotreating organic fats and fatty acids are distinct to the usual reactions

taking place in a refinery hydrotreater. Refinery hydrotreaters are designed to remove sulfur from

petroleum fuels. This process is known as hydrodesulfurization (HDS) and, as discussed earlier, is used to

reduce the sulfur content of finished fuels in order to meet increasingly stringent fuel specifications such

as required for Ultra Low Sulfur Diesel (ULSD, <10 ppm S). Sulfur emission regulations are tightening

around the world and, consequently, hydrogen consumption in oil refineries is projected to double over

the next decade. While renewable oils typically do not contain much sulfur and thus do not require

hydrodesulfurization, hydrotreatment in the form of hydrodeoxygenation (HDO) is still needed to

remove oxygen. Unfortunately, hydrodeoxygenation of renewable oils requires more hydrogen gas

inputs than do hydrodesulfurization of crude oils. For 100% renewable feed, a hydrogen consumption of

300-400 Nm3/m3 is not unusual (Egeberg et al., 2010); compare this with about 34 Nm3/m3 for the

hydrotreating of 1% sulfur petroleum (Stratiev et al., 2009). The presence of oxygen in the feed also

increases reactivity and results in the formation of byproducts such as propane, water, carbon monoxide

and methane (Egeberg et al., 2010). These gases must be removed by increasing the gas purge rate in

the system. If not removed, these gases will cause numerous problems such as: a) altering the hydrogen

partial pressure or reducing catalyst activity: b) CO and CO2 competing with S- and N- species for

hydrotreating catalyst sites; and c) liquid water and CO2 reacting to form corrosive carbonic acid in the

effluent train of the reactor. The formation of these carbonaceous byproduct gases diverts carbon from

the final fuel and thus reduces process yields compared to fossil diesel hydrotreating (Egeberg et al.,

2010). Methane in particular is a highly undesirable byproduct because it not only diverts one mole of

carbon from the fuel, it also diverts four moles of elemental hydrogen (two moles of molecular

hydrogen, H2) from the reaction mass thus unproductively increasing hydrogen consumption in the

process. These challenges have been documented by ConocoPhillips in trials performed in refineries in

Texas, USA and in Ireland (ConocoPhillips, 2010).

Due to all of the above challenges, renewable oils have not yet been processed in a neat (100%) form in

conventional refineries. The few trials of co-processing vegetable oils with petroleum liquids that have

been carried out in commercial refineries have used low percentages of renewable oils and have only


July 2014 59

been partially successful. Poor desulfurization, hydrogen starvation and pressure drop build-up are

among the issues that have been encountered using reactor setups and catalysts not specifically

designed for renewable oil feeds. As an example, Haldor Topsoe, a major refinery catalyst producer,

reported an industrial trial of co-processing a few percent vegetable oil in a ULSD hydrotreating facility

(Egeberg et al., 2010). A few days after the renewable oil feed was introduced to the ULSD hydrotreater,

the pressure drop-in the reactor increased such that the refinery had difficulty continuing operation. In

related work, Haldor Topsoe showed that the CO byproduct from co-processing this biofeed reduced

hydrodesulfurizing (HDS) and hydrodenitrogenising (HDN) activities of conventional CoMo-type

catalysts. As NiMo catalysts did not seem to be affected by blend levels up to 15% biofeed, these type

catalysts are currently favoured when designing hydrotreatment catalysts for renewable oil co-

processing (Egeberg et al., 2010).

Another technical challenge for designing effective catalysts for biofeed processing is that the catalyst

needs to have some dewaxing activity to enable normal paraffinic molecules to be cracked and

isomerized to lighter molecules thereby improving the cold flow properties of the final fuel or

blendstock product. The dewaxing requirements for catalysts processing biofeed are higher than those

processing petroleum feed. The long and largely unbranched acyl chains in vegetable oils yield paraffinic

chains of similar length and level of isomerisation (branching). In contrast, petroleum-derived feeds to

hydrotreaters, such as light gas oil, are typically already more isomerized and cracked (partly due to the

distillation and cracking processes they have been through prior to arriving at the hydrotreating unit).

Dewaxing is only essential if the feed has a high content of renewable oil as at low concentrations the

cold flow issues are not as prominent and they can be alleviated by blending light fractions and cloud

point-depressing additives.

This complex challenge of designing catalyst beds that can help perform all the aforementioned

selective reactions is currently being tackled by companies such as Haldor Topsoe, UOP Honeywell and

other catalyst innovation companies. The goal is to improve oil refining catalysts and process designs for

the purpose of improving the ability to process biofeeds in hydrotreating reactors.


July 2014 60

It is important to note that, as well as addressing the technical challenges of co-processing oleochemical

based biofuels in conventional oil refineries, regulatory hurdles must also be resolved. For example, the

current US Renewable Fuel Standard (RFS) mandates definition of “biomass-based diesel” does not

allow any renewable fuel derived from biomass to be co-processed with petroleum feedstocks. This co-

processing exclusion mostly affects those biofuel intermediates which have the greatest opportunity for

refinery leveraging such as vegetable oils. (Other technologies such as pyrolysis are less affected as,

unlike oleochemical fuels, they also qualify under the other, mandated “advanced biofuels” categories,

of the RFS). It has been suggested that, for co-processed vegetable oils in particular, this RFS limitation

translates to a competitive disadvantage of up to $2/gal ($7.6/L) compared to conventional FAME

(Weyen, 2012).

2.4 Commercialization aspects

2.4.1 Feedstock sensitivity

The chain length of the fatty acids in the TAG feedstock determines the products of a HEFA facility. Most

of the feedstocks available today are derived from vegetable oils. These typically contain long fatty acid

Box 2-1: SunPine: Deriving diesel from tall oil: A renewable fuel from the forest

In 2009, Preem AB, a Swedish oil refining company, partnered with Sunpine AB, a producer of tall oil

from Kraft paper mills, to produce green diesel (www.preem.se). Tall oils are a byproduct of pulp

mills and they are mainly derived from the resins and extractives present in softwood feedstocks

such as pine, spruce and birch. Due to being primarily comprised of large amounts of resin acids and

free fatty acids they are very acidic. To improve transport logistics tall oils are typically processed to

FAMEs prior to shipping from the pulp mill. The resulting tall oil FAME liquids are known by the

abbreviation RTD (raw tall diesel). Under this collaboration, the Preem AB oil refinery in

Gothenburgh, Sweden developed and demonstrated the ability to hydrotreat up to 30% RTD blends

with 70% mineral oil, which is a record high biofeed fraction for co-processing in a conventional

refinery (Egeberg et al., 2010). Haldor Topsoe provided the catalyst beds and process design. It

should be noted that hydrotreating FAMEs (such as RTD) is distinct from hydrotreating vegetable oil

TAGs with the major difference being that a high yield of byproduct methane is obtained processing

FAMEs as opposed to propane processing TAGs.

http://www.preem.se/


July 2014 61

chains corresponding to the carbon chain length of diesel i.e. C16-C22. These molecules can be cracked to

shorter chains to fit the “lighter” jet fuel and gasoline range. However, the cracking step is not

sufficiently selective and creates by-products and reduces the overall fuel yield. For example, when a

long alkyl chain is cracked, only some of the chains are of the desirable length while a number of

undesirable short chains are also produced. These “too-short-waste” chains form a light naphtha-like

byproduct and the overall fuel yield is reduced. Although the oils from camelina, palm kernels and most

cyanophyta contain TAGs with shorter chain fatty acids (which are in the jet fuel range) (Bauen, Howes,

Bertuccioli, & Chudziak, 2009; Pearlson, 2011) these feedstock’s are currently only available in relatively

small volumes. Camelina and cyanophyta oils are only produced in small volumes while, in year 2011/12,

only about 6 million metric tonnes of palm kernel oil (not to be confused with palm oil) are produced

annually of the global 160 million metric tonnes of vegetable oils production (USDA, 2013).

2.4.2 Feedstock procurement

While there are fewer technological barriers that have to be resolved to achieve techno-economically

effective processing of lipids to finished drop-in fuels, lipid feedstocks themselves are likely to be

difficult to source cheaply and sustainably in the volumes required for significant production of biofuels.

The most widespread source of lipids for biofuel production are vegetable oils such as rapeseed and

palm oils which are currently used extensively in the food market. Thus there is competition with regard

to price and access to prime agricultural land for these feedstocks. Moreover, the price of these

feedstocks can be and often is higher than the price finished diesel fuel commands. Commodity prices

for food-grade vegetable oils are generally higher than diesel prices. For example, on January 15, 2013,

rapeseed oil (also known as canola oil), a popular food-oil as well as a common feedstock for biodiesel

production, had an average commodity price of $4.21 USD/gal ($1.11/L) while diesel fuel for the same

month had a price of $3.22 USD/gal ($0.85/L) (IndexMundi, 2013). Oleochemical feedstock prices also

appear to be linked to petroleum prices as depicted in Figure 2-4, which shows this for palm oil, one of

the lower cost HEFA feedstocks. The combined production potential of HEFA is estimated to currently be

in the low hundreds of thousands of barrels per day and there are significant constraints on increasing

near-to-midterm production capacity, especially since these facilities would be competing for the same

feedstock as existing biodiesel producers (Hileman et al., 2009). Global production of vegetable oils is

currently about 3 million barrels per day. A thirty-fold (30x) increase in vegetable oil production would

be required to satisfy the current 44 million barrels per day of global transportation fuel demand (IEA,

2012b)


July 2014 62

The likely ongoing high cost of feedstock is a major impediment to successfully implementing and

expanding oleochemical platforms. For example, a recent report by MIT’s research program “PARTNER”

assessed the cost to US commercial aviation of meeting the US Federal Aviation Administration’s (FAA)

aviation biofuel goal to use 1 billion gallons (3.8 billion L) of renewable jet fuel per year from 2018

onwards (Hileman et al., 2009). This analysis only considered the production and distribution costs of

HEFA derived biojet fuel (“biojet”). One of the conclusions is that biojet will require a premium (implying

the need for a subsidy) of $2.69 USD/gal ($0.71/L) compared to petroleum-derived jet fuels. It was

suggested that this premium would have to be voluntarily paid by the aviation (and military) sector if

they wanted to reach their FAA goal. It should be noted that this premium would not drop significantly

by including aviation fuels in the RFS mandate. The only scenario where the premium is reduced

significantly, down to $0.35 USD/gal ($0.09/L), is the case where fallow rotation land (land planted with

oilseeds in between corn and oil seed crop cycles) is sufficiently available to produce all the vegetable oil

feedstock needed to meet the FAA’s goal. This is unlikely due to the generally low productivity of fallow

land and the relatively low yield of possible new oilseed crops such as camelina.

Figure 2-4: Commodity prices of Diesel and Palm Oil

Source: (IndexMundi, 2013)

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

Oct-06 May-07 Jan-08 Aug-08 Apr-09 Dec-09 Jul-10 Mar-11 Oct-11 Jun-12 Jan-13

USD

/gal

Diesel Price (US Dollars per Gallon)

Palm oil Price (US Dollars per gallon)


July 2014 63

However, various government and industry funded R&D programs are currently trying to improve the

productivity of oilseed crops that are suitable for biofuel and for biojet applications in particular. One of

the overall goals is to develop oilseed crops that compete less with food markets by producing non-

edible oils with lipid fatty acid composition favourable for biofuels and growing these crops on “marginal

land”. The USDAs “Farm to Fly” program is a good example of this type of strategy (USDA, 2012).

Alternative oleochemical feedstock sources which do not use arable land include algae (autotrophic), tall

oils (a pulp mill residue as seen in Box 2-1), and waste oils and fats.

Algae have also been suggested as an alternative, non-land-use-intensive source of lipids. These micro-

organisms are able to capture CO2 and sunlight and produce lipids without utilizing productive arable

land. However, major issues such as yields, maintenance of favourable growth conditions in large scale

ponds and extraction of lipids in a usable form have, so far, limited commercialization. The potential of

algae biofuels is more extensively reviewed in a previous IEA Bioenergy Task 39 report (Darzins et al.,

2010).

As well as autotrophic algae, which use CO2 as their carbon source, heterotrophic algae using sugar as

their carbon source have been used to produce TAGs and fatty acids for drop-in biofuel production.

While autotrophic algae require optimal exposure to sunlight, heterotrophic algae do not and thereby

avoid some of the operational challenges of operating raceway ponds or photobioreactors. However,

using heterotrophic algae requires securing a cheap and sustainable source of sugar feedstock and

proving that this mode of algal production can be scaled up economically enough to enable profitable

biofuel applications.

Waste oils (e.g., tall oils) and used cooking oils (UCO) have been used as feedstocks for drop-in biofuels.

One benefit with these types of feedstocks is that, compared to purpose grown oils, the carbon

footprint of used oils has already been absorbed by the life cycle of another product or service.

However, UCOs are typically only available in small quantities and at dispersed locations (e.g.,

restaurants) so their collection and quality control is challenging. Although industrial waste oils such as

tall oils are more centrally accumulated and there are facilities using these oils commercially today such

as the Sunpine facility in Sweden (see Table 2-2) there are again limited volumes of this material

available globally.

http://task39.org/files/2013/05/IEA-Task-39-Current-Status-and-Potential-of-Algal-biofuels.pdf


July 2014 64

2.4.3 Commercial facilities

As mentioned earlier, oleochemical derived fuels are the only drop-in biofuels that are being produced

at relatively large commercial scale today. However, as shown in Table 2-2, they account for less than

one hundred thousand barrels of fuel per day. Neste Oil, a Finnish petroleum refining company, is

currently the world’s largest producer of drop-in biofuels and operates 3 HEFA facilities (Finland,

Rotterdam, Singapore) with an annual total capacity of 630 million gallons (2.4 billion L) of palm oil-

derived diesel, which is marketed as “NexBtL” (Neste Oil, 2013a). In Q1 of 2013, the company’s

renewable fuel division recorded an operating profit of 26 million euros (Neste Oil, 2013b). Other

commercial HEFA manufacturers include three facilities in southeastern USA. One is the joint venture

between Syntroleum and Tyson foods, which licensed their “Biosynfining” technology to a Dynamic

Fuels commercial plant in Louisiana that currently produces 75 million gallons (284 million L) per year of

green diesel (renewable diesel). Another, Honeywell-UOP, which licensed their technology to the

Diamond Green Diesel facility in Kentucky (a joint venture between Darling International Inc. and Valero

Corporation) for a 136 million gallon (515 million L) facility in Norco, Louisiana. The third is Emerald

biofuels, which announced in May 2012 that it will build an 85 million gallon (322 million L) per year

capacity plant at the Dow Chemical site in Plaquemine, Louisiana. In the EU, the ConocoPhillips

Whitegate refinery in Cork, Ireland produces 1000 barrels per day of HEFA by co-processing soy oils with

petroleum.

Table 2-2: Current world annual production capacity of HEFA drop-in biofuels

Company Feedstock million gallons/yr

million L/yr

barrels per day

Source

Neste Oil Palm oil 626 2,371* 45201 (Neste Oil, 2013a)

Diamond Green Diesel Tallow 136 515 10000 (Diamond Green Diesel, 2013)

Emerald Biofuels Tallow 85 322 6133 (Emerald Biofuels, 2013)

Dynamic fuels Tallow 75 284 5411 (Dynamic Fuels, 2013)

Conoco Phillips Whitegate Refinery

Soy oil 13.9 52 1000 (Conocophillips, 2013)

Sun Pine Tall oil 26 100 1906 (Chemrec, 2009)

World Total 963 3644 69651

Bold figures are from source and all other figures calculated (bbl/day calculations based on 330 day/year operations) *Neste Oil source listed 1,980,000 metric tonnes of diesel which were converted to 2,371 m L using diesel density of 0.832 kg/L

http://www.conocophillips.com/EN/about/company_reports/spirit_mag/Pages/whitegate_story.aspx


July 2014 65

Other European projects have been announced by ENI and Galp Energia, which both intend to produce

green diesel (Maniatis et al., 2011), although neither project has as yet started construction. As

mentioned earlier, pulp and paper companies such as Sunpine and UPM Kymene produce HEFA using

their tall oil (2% of wood feedstock) as feedstock. While the 100,000 t/year, 150 million euros UPM

Lappeenranta facility (Kaukas mill) in Finland is under construction, Sunpine’s facility in Piteå, Sweden

has been operating since 2007.

2.4.4 Fuel Quality

The process of hydrotreating vegetable oils can result in the production of high quality HEFA fuels that

exceed the specifications of petroleum based transportation fuels (Table 2-3). As an example, HEFA

derived diesel and jet fuel have essentially no sulfur content whereas their petroleum counterparts can

contain up to 3000 ppm of sulfur. Other improved characteristics of HEFA fuels include higher energy

density, lower aromatics content and for diesel HEFAs higher cetane number (Table 2-3).

Table 2-3: Selected properties and specifications of fossil and renewable HEFA diesel and jet fuels.

Property Diesel Jet

Fossil 1HEFA Fossil 2HEFA

Oxygen content wt % 0 0 0 0

Specific gravity kg/L 0.84 0.78 0.75-0.84 0.73-0.77

Cetane 40-52 70-90 - -

Sulphur ppm <10 <2 <3000 <15

Specific energy MJ/kg 43 44 >42.8 44.1

(typical)

Aromatics Vol % <12 0 <25 <0.5

1Properties of renewable diesel from UOP Green Diesel. 2ASTM D7566 Annex 1 used for hydroprocessed renewable oil specification. Source: (Pearlson, 2011) The absence of aromatics in renewable HEFAs is generally viewed as an advantage from an air pollution

standpoint since phenolic compounds are associated with emissions of polyaromatic hydrocarbon (PAH)

pollutants (European Commission, 2001). While aromatics are generally undesirable in petroleum fuels,

a minimum amount is actually necessary to meet transportation fuel specifications. Aromatics are

energy dense molecules and are responsible for the necessary swelling of seal elastomers in an engine’s

fuel system. The absence of aromatics in HEFA means that these “drop-in” biofuels will be blendstocks


July 2014 66

that need to be blended with petroleum jet fuel. This is part of the reason why the ASTM standards have

only approved 50% blends of HEFA biofuels for jet use (Bauen et al., 2009; Hileman et al., 2009).

2.4.5 Test flights and certification

As mentioned earlier, the aviation industry is uniquely dependent on drop-in biofuels as the only real

alternative to current petroleum-derived jet fuels. Conventional biofuels such as ethanol and biodiesel

are not suitable for jet engines. The unique dependence of aviation on drop-in biofuels is one of the

primary drivers for ongoing commercialisation efforts. To date, the vast majority of all biofuel test flights

have been based on oloechemically derived HEFAs. As shown in Figure 2-5, Virgin Atlantic conducted

one of the earliest biofuel test flights in 2008, and a number of other commercial airlines and the US

Navy have successfully demonstrated biofuels for aviation applications since then. These efforts

contributed to ASTM’s final approval of 50% HEFA blends in aviation fuels in July 2011. Note that 100%

HEFA has not yet been tested commercially and the only flight that has been performed on pure biojet

was an experimental flight by the Canadian National Research Council in October 2012. This flight was

also one of the world’s first to use biojet made from hydrotreated oils derived from an Ethiopian

mustard variety as commercialized by the Canadian company Agrisoma (www.agrisoma.com).

As illustrated in Figure 2-5, a gasification-derived Fischer-Tropsch (FT) biojet fuel was approved 2 years

prior to 50% HEFA without any prior test flights or these gasification-derived biofuels being

commercially available. The main reason behind this rapid approval was because of prior certification of

coal derived FT jet fuels. Sasol’s semisynthetic jet fuel blends (containing 50% coal-derived FT-Jet and

50% petroleum-derived jet) were approved by ASTM for use in aircrafts in 2009, after a 7 year

certification process. The certification of biomass FT-based jet fuels was justified on the grounds of the

chemical equivalence between purified biomass syngas and coal syngas. Given their chemical

equivalence, the functional equivalence was assumed by the ASTM and consequently no further testing

was requested. In contrast, HEFA fuels have no chemical equivalence to any prior certified

transportation fuel and hence their approval by ASTM took more time and testing in order to provide all

the assurances of functional equivalence. Jet fuels have one of the most stringent ASTM specifications.

While the alcohol-to-jet (ATJ) technologies will be discussed in Chapter 5, it should be noted that, as

shown in Figure 2-5, ASTM certification for ATJ aviation fuels is expected to be approved in early 2014.

http://www.agrisoma.com/


July 2014 67

Figure 2-5: Timeline of biofuel test flights and ASTM certification approvals Source:(Alexander, 2012; ATAG, 2011; IATA, 2013; NRC Canada, 2012; SAFUG, 2014)

While ASTM certification is already in place for HEFA jet fuels, the cost and sustainability (i.e.,

availability) of the feedstock remain major challenges constraining extensive HEFA commercialization.

2.4.6 Sustainability certification of HEFA

Sustainability certification of oleochemical routes to drop-in biofuels is an ongoing concern,

predominantly affecting the HEFA drop-in biofuel platform’s feedstock sourcing. If the possible

emissions due to any land-use change are ignored, the life-cycle GHG emissions of HEFA are estimated

to be around half those of petroleum jet fuels (Hileman et al., 2009). However, sourcing vegetable oil

feedstock for HEFA facilities will likely mean growing oilseed crops on land that will displace natural

habitats or that could otherwise be used for food production. This production system has been subject

to public criticism on the grounds of land use change. Examples include public concern regarding Neste

2008 2009 2010 2011

Late 2009: ASTM

approval

of 50% blend FT

2012 2013

Feb 2008: 1st

commercial

test flight

June 2011: 1st

commercial flight

with passengers

2014

(expected):

ASTM approval

of 50% blend

ATJ

Oct 2012: Canada

NRC: 1st test flight

on 100% biofuel

2014

Various test flights by: Air NZ, Lufthansa, Continental, Japan Airlines, US Navy, US Air Force, TAM,

Honeywell, Air China, Alaska, Etihad, JAL, TAM, Interjet and others (see:

http://legacy.icao.int/icao/en/env2010/ClimateChange/GFAAF/Summary.htm)

July 2011: ASTM

approval

of 50% blend

HEFA

http://legacy.icao.int/icao/en/env2010/ClimateChange/GFAAF/Summary.htm


July 2014 68

Oil’s (Finland) aspirations for palm plantations in Malaysia and the Friends of The Earth report entitled,

“Take-off in the Wrong Direction” which focused on detrimental land use change in Java, Indonesia, due

to palm oil demand for jet fuel flights in Europe (FOE, 2012). While land use changes are difficult to

quantify (Finkbeiner, 2013), various certification schemes such as the “Roundtable on Sustainable

Biofuels” are currently considering the inclusion of indirect land use change (ILUC) in their set of

standards (RSB, 2012). The company SkyNRG (a KLM – NSGSA collaboration) recently earned RSB

certification for its entire supply chain and is currently the only fuel operator in the world that can

deliver certified renewable jet fuel at any airport.

In summary, the oleochemical platform is well positioned to be the “first generation” approach to drop-

in biofuel production due to the low oxygen content and high H/C ratio of oleochemical feedstocks. This

platform for drop-in biofuels is already producing HEFAs at commercial scale with a total global capacity

of about 70,000 barrels per day. Although this represents a small fraction of total global transportation

fuel demand (44,000,000 barrels per day, 5,000,000 of which are jet fuels, IEA (2012)), it is the only

commercially available drop-in biofuel that has been produced at significant volumes to date. The main

challenges to further development of this platform mainly relate to feedstock availability, cost and

sustainability. Lipid feedstocks are relatively scarce and expensive and they come with potential

sustainability challenges such as land use change and competition with food markets. Prices for lipid

feedstock have historically tracked with petroleum prices and they have also been priced higher than

diesel fuel for long periods of time. It is clear that if HEFAs continue to be the only commercially

available “biojet” fuel option, fulfilling aviation biofuel targets will likely result in higher operating

expense for airlines. For example, in order to meet the US FAA target of 1 billion gallons (3.8 billion L) by

2018, it has been estimated that US aviation stakeholders would have to pay a premium of about $2.7

USD/gal ($0.71/L) for HEFA derived jet fuel. However, the advantages of using biofuels in aviation are

well established and, if the sustainability concerns can be resolved, this close relationship between HEFA

derived drop-in biofuels and aviation is likely to continue to grow.


July 2014 69

CHAPTER 3: THE THERMOCHEMICAL PLATFORM

Thermochemical processes use high temperatures and catalysts to convert biomass to liquid biofuels

and chemicals as well as heat and power (Brown, 2011). Unlike oleochemical technologies, which often

use lipid feedstocks, these processes typically use lignocellulosic biomass as the feedstock. The biomass

is reacted at high temperatures (> 500 °C) to form carbonaceous gases and liquids as well as char solids.

The two main thermochemical routes are gasification and pyrolysis. The gasification process, as the

name implies, converts biomass mainly to a gaseous intermediate, known as syngas. The pyrolysis

process, on the other hand, maximizes the production of pyrolysis liquids, also known as pyrolysis oils or

bio-oils. The gaseous and liquid intermediates of these thermochemical processes are mostly comprised

of oxygenated species and thus need to be further processed to produce drop-in blendstocks. Using the

Fischer-Tropsch catalytic process, syngas can be catalytically condensed to form liquid hydrocarbon

mixtures known as FT liquids that, in turn, can be upgraded to fuels for gasoline, diesel and jet engines.

Similarly, pyrolysis oils can be upgraded to liquid transportation fuels after further processing using

catalysts and hydrogen. The main objective of catalytic upgrading is to remove the oxygen from both the

syngas and bio-oil derived intermediates in order to produce petroleum-like hydrocarbon fuels. This

deoxygenation process requires a chemical reducing power which is typically supplied by hydrogen

derived from natural gas. As mentioned previously, the biomass itself can be used as a source of

renewable hydrogen but this will result in a significant drop-in overall process yields. Bio-oil upgrading

processes are usually conducted in relatively complex facilities that require both high hydrogen inputs

and capital costs (Bridgwater, 2012). In most thermochemical processes there is a trade-off between

capital costs, product yield and the extent of hydrogen requirements.

Co-locating thermochemical processes at refineries can be used to leverage oil refinery assets, reduce

capital costs and ensure a relatively low cost source of hydrogen. Pyrolysis platforms appear particularly

well suited to exploit co-location and synergy with existing refineries as pyrolysis oils can be processed

using similar equipment to that currently used to upgrade crude oil. However, in practice, pyrolysis

liquids contain relatively high levels of water and oxygenated species and thus are chemically quite

distinct from crude oil and poorly suited to being “dropped into” existing petroleum processing units.

However, it is likely that downstream refinery units such as FCCs and hydrocrackers could be configured

to process thermochemical biofuel intermediates such as FT liquids and hydrotreated bio-oils to drop-in

fuel blendstocks.


July 2014 70

3.1 Overview of thermochemical processes

The most basic and widely applied thermochemical process is direct combustion of lignocellulosic

biomass to produce heat and electric power. Combustion, a process recognised since the dawn of

humanity, accounts for the vast majority of bioenergy applications in the world today. From relatively

primitive open cooking fires and charcoal production through high efficiency industrial boilers and

district heating systems, biomass combustion represented more than half of global renewable energy

production in 2010 (IEA, 2012b).

From a technology standpoint, the combustion process is relatively simple and well understood (R. C.

Brown, 2011). It entails the rapid reaction of biomass fuel with excess oxygen to generate thermal

energy as well as highly oxidized flue gases, mainly CO2 and H2O. The chemical energy in the biomass is

converted to thermal energy and, under optimized conditions, the exothermic reaction almost

completely oxidizes the biomass. The temperatures of the generated flames can exceed 1650 °C (R. C.

Brown, 2011).

Direct combustion of biomass can be used to indirectly power electrified transportation fleets. From

purely a GHG emission savings perspective, biomass-powered electric vehicles can be superior to

biofuel-powered internal combustion engine vehicles (Campbell et al., 2009). However, as has been

discussed elsewhere, electric vehicles are relatively expensive, still require improved battery

technologies and are mostly limited to light duty and short haul transportation applications (IEA, 2012).

Whereas combustion requires molecular oxygen (O2) to be highly effective, charcoal production involves

the “burning” of biomass in the presence of limited oxygen. Charcoal production can be considered the

predecessor of pyrolysis and gasification processes. In its primitive form, charcoal making is conducted

in clay-covered wood piles with a flue opening in the middle. A wood fire is started at the bottom of the

flue and it slowly smolders the covered wood over a couple of days. Although modern industry uses

more advanced charcoal production processes, this ancient technique is still widely practiced in less

industrialized global communities, typically yielding about 60% by volume (25% by mass) of charcoal

from the original biomass. Along with solid charcoal, this process produces liquid tar as well as flue

gases.


July 2014 71

Transportation applications of small biomass gasifiers for vehicles were developed during WWII due to

reduced availability of petroleum in portions of the world, such as Scandinavia. Since the 1970s oil

crises, gasification of biomass has received considerable research attention as potential sources of

renewable liquid and gaseous fuels with subsequent development of fast pyrolysis for liquid fuel

production since the 1980s. Unlike in traditional pyrolysis process for charcoal making, where the target

product is the solid char, in the new fast pyrolysis the target product is liquid (bio-oil) and in gasification

it is synthesis gases (syngas). By adjusting the processing conditions, pyrolysis can maximize the

proportion of liquid products and gasification can maximize the proportion of gases. As shown in Figure

3-1, pyrolysis is conducted at intermediate temperatures of about 500 °C, in the absence of oxygen, and

it produces a mixture of gases, char and liquids (water and water soluble and water insoluble organics).

However, in fast pyrolysis, the residence time is reduced to a couple of seconds or less and the

proportion of liquid yields can reach as high as about 75% by mass. In gasification, the biomass is

reacted under pressures of 1-40 bar and at temperatures exceeding 800 °C and in the presence of

regulated amounts of oxygen. Under these conditions the production of gases is favoured and can reach

up to 85% by mass of the total products (Bridgwater, 2012).

Figure 3-1: Product spectrum from thermochemical conversion of biomass Source: Bridgwater 2012

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Fast pyrolysis (2 sec @500 C)

Slow Pyrolysis (longresidence)

Gasification (few sec @ >800 C)

organics (liquid)

water

char

gas


July 2014 72

Both syngas and bio-oil are fluid biomass intermediates that can be used as combustion fuels for

stationary power applications such as burners, boilers, furnaces and industrial kilns. However, for drop-

in biofuel applications, these intermediates need to be catalytically upgraded to oxygen-free

hydrocarbons as shown in Figure 3-2. This upgrading takes various forms such as Fischer-Tropsch (FT)

condensation to produce paraffins and 2-stage hydrotreatment to produce hydrotreated pyrolysis oils

(HPO). To maximize yields, both upgrading technologies use specialized catalysts and hydrogen inputs.

The resulting FT liquids and HPOs are both hydrocarbon mixtures that need to be subsequently distilled

and hydrocracked in order to produce a mixture of gasoline, jet and diesel range hydrocarbons. In a

similar fashion to the hydrocracking of vegetable oils (Chapter 2) and, depending on how the

hydrocracking process is conducted, the proportion of gasoline, diesel and jet fractions can be adjusted.

Figure 3-2: Simplified representation of major thermochemical drop-in biofuel process routes

Although pyrolysis and gasification have many fundamental characteristics in common, the two

processes differ markedly in the details of their associated biomass intermediates upgrading

technologies, drop-in fuel yields, capital costs and hydrogen (H2) requirements.

PRODUCT INTERMEDIATES CATALYTIC

UPGRADING

Hydro

treatm

ent

1

Pyrolysis oil

Fis

cher-

Tro

psc

h Gasification

Syngas

Biomass

Hydro

cra

ckin

g

HPO

FT liquids

Gasoline

Diesel

Jet

Gases

500°C No O

2

900°C some O

2

Hydro

treatm

ent

2


July 2014 73

3.2 Fast Pyrolysis

As noted earlier, fast pyrolysis is the thermal processing of solid biomass in the absence of added oxygen

to produce bio-oil, which can be considered as an intermediate towards the production of drop-in

biofuels. The production and properties of bio-oil and the technologies that can upgrade this biomass

derived liquid to transport fuels as well as some techno-economic aspects of fast pyrolysis biofuels are

discussed in the next section.

3.2.1 Bio-oil production

Fast pyrolysis is a thermal decomposition process which requires rapid heating of biomass to about 500

°C and a subsequent rapid cooling of the resulting vapours to room temperature. Upon cooling, these

vapours condense to form the liquid bio-oil product. It has been demonstrated that rapid heating and

cooling is crucial to maximizing bio-oil liquid yields at the expense of char and gas production

(Bridgwater, 2012). Slow heating favours the production of solids such as charcoal while a long

residence time in the high temperature zone of the reactor results in further cracking of vapours which

consequently favours the production of permanent gases. A long residence time can also promote the

polymerization of vapour molecules and the formation of solids. To maximize bio-oil yields (to about

75% of starting biomass by mass) rapid heating to the target temperature must be achieved throughout

each biomass particle (i.e., within about one second). These high heat transfer rates (up to 1000 °C/s)

ensure maximum devolatilization (vaporization) of the biomass solids and, so far, have only been

achieved by a select number of reactor designs (Bridgwater, 2012).

One reactor type which is well suited for fast pyrolysis is the Bubbling Fluidized Bed (BFB) reactor. This

reactor uses a hot sand fluidized bed to achieve high rates of heat transfer to biomass particles. The

reactor beds are fluidized using a compressed carrier gas which is fed through the bottom of the reactor

at sufficiently high rates to “fluidize” the loose solids contents of the reactor (sand and biomass) while

transferring the gas-entrained char upwards. These types of BFB reactors have been used by the

petroleum industry for the gasification of coke since the 1950s. They are robust systems that achieve

high heat transfer rates and uniform bed temperatures (Ringer et al., 2006) which are both highly

desirable attributes for fast pyrolysis reactions.

As depicted in the simplified schematic of a typical BFB fast pyrolysis process (Figure 3-3) the biomass is

first dried and ground to a particle size of about 3 mm to facilitate rapid particle heat up and

devolatilization. These particles enter the fluidized bed pyrolysis reactor where they are rapidly heated


July 2014 74

to a temperature of about 500 °C. After about a 2 second residence time, the generated vapours are

vented to a cyclone where they are separated from the entrained solid char particles. The recovered

char can be sold as a value-added product (soil amendments and activated carbon, e.g., as described in

Dynamotive’s business model) or used as fuel for the furnaces that generate and compress the hot

recycle gas that feeds the main pyrolyser reactor. The clean vapours are then swiftly transferred to a

quench cooler where they are condensed to form the bio-oil. The uncondensed fraction of the vapours

along with the permanent gases is then transferred to a second condensation train such as a coalescing

filter, scrubber or electrostatic precipitator (e.g. Nexterra) where additional bio-oil is recovered. The

remaining flue gas is fed to the furnace that generates hot gas for the main reactor. Fast pyrolysis oils

contain up to 75% of the mass and 65% of the energy that was contained in the original biomass

feedstock (Bridgwater, 2012). These types of BFB reactors have been used by the Canadian company

Dynamotive at a semi-commercial scale (see Table 3-1) as well as at a smaller, demonstration scale such

as the 200 kg/hr unit of Union Fenosa in Spain. Both the Dynamotive and the Union Fenosa facilities are

based on a design developed at the University of Waterloo and commercialized through its Canadian

spin-off company RTI (Resource Transformations International) (Bridgwater, 2012; Czernik & Bridgwater,

2004).

Figure 3-3: Simplified schematic of bubbling fluidized bed (BFB) fast pyrolysis Source: adapted from Bridgwater, 2012

Fluid bed

reactor

500oC, 2 sec,

atmospheric

pressure 3 mm

biomass particles

Volatiles and char

cycl

on

e

Recycle

gas

heater

Ch

ar

Volatiles (char-free)

Qu

ench

co

ole

r

Elec

tro

stat

ic

pre

cip

itat

or

Pyrolysis liquids

75 wt% yield

Gas

Gas

Condensate


July 2014 75

A more complex version of the bubbling fluidized bubbling bed (BFB) reactor is the circulating fluidized

bed (CFB) reactor. This reactor configuration has been used by the petroleum industry for many decades

and it has a long history of industrial operations especially in the fluidized catalytic cracking (FCC) units

described earlier. This type of system (Figure 3-4) is similar to the BFB process except the compressed

recycle gas is fed at much higher velocities, such that the entire loose contents of the reactor (vapors,

gases, char as well as the fluidized bed’s sand particles) are carried into the downstream cyclone. The

char and sand are then recovered from the cyclone and they are then fed together to a combustor,

where the char is burned off to heat up the sand. The cleaned hot sand (at about 800 °C) is then fed

back to the main reactor entrained in the compressed carrier gas and the process cycle is repeated. This

system is more expensive to install and operate than the BFB process but it comes with the advantages

of constantly regenerating clean sand bed particles and achieving higher throughputs. CFB requires

careful sizing of the biomass particles since the rapid gas flow only permits a very short residence time in

the hot zone of the pyrolysis reactor. The CFB pyrolysis is the configuration of choice for the Canadian

pyrolysis company “Ensyn” and it is marketed under the name RTP (Rapid Thermal Processing). Other

developers of the CFB configuration include CRES (catalytic pyrolysis, Greece) and Ensyn for ENEL (Italy)

(Ringer et al., 2006) and VTT-led consortium in Finland (Metso, UPM, Fortum).

Figure 3-4: Simplified schematic of circulated fluid bed (CFB) fast pyrolysis Source: adapted from Bridgwater, 2012

Ch

ar & san

d

Volatiles char & sand

Fluid bed

reactor

500oC, 2 sec,

atmospheric

pressure

3 mm Biomass particles

Cyc

lon

e

Co

mb

ust

or

Volatiles (char-free)

Qu

ench

co

ole

r

Elec

tro

stat

ic

pre

cip

itat

or

Pyrolysis liquids

75% yield

Gas

Gas

Condensate


July 2014 76

The main drawbacks of fluidized bed reactors is that they rely on a compressed carrier gas which often

carries char contaminants to the bio-oil product and compressing this gas also requires high capital and

operating costs. The carrier gas, used in order to mix and circulate the sand bed, carries char particles of

such small submicron size that even the solids separation cyclone cannot capture them. Thus, these tiny

particles remain entrained in the vapour that enters the quench cooler and they end up in the bio-oil

product (Bridgwater, 2012). This can be a problem as char particles can catalyze tar and coke formation

and they can plug reactor pipes and filters upon subsequent bio-oil upgrading. It should also be noted

that the compressors used to deliver high speed carrier gases are capital-intensive and they are not well

suited for small scale applications (Wright et al., 2010).

Alternative reactors that do not use a carrier gas have recently been developed. These reactors use

centrifugal forces and mechanical motion to achieve the high rates of heat transfer needed to rapidly

volatilize the biomass particles. These types of pyrolysis reactors include ablative and rotating cone

designs based on the principle of sliding biomass particles against a hot surface, thus “melting” the fibre

in a similar way to a block of butter melts when pressed against a hot surface. Ablative pyrolysis reactors

do not use a fluidized bed or sand particles while rotating cone reactors use sand particles contacting

biomass particles but without using fluidization by a carrier gas (Bridgwater, 2012; Venderbosch & Prins,

2011). The concept of ablative pyrolysis was first proposed by the CNRS laboratories in Nancy, France.

Subsequent ablative reactor designs have been developed by NREL in the USA (vortex reactor) and by

Aston University in the UK (plate reactor) (Bridgwater, 2012). The company Pytec has a demonstration

plant for ablative pyrolysis in Germany. The rotating cones reactor concept was initially developed by

the Dutch Company BTG (a University of Twente spin-off) which currently operates a 5 tpd

demonstration facility in the Netherlands. BTG had designed and built a 48 tpd facility which was

operated in Malaysia several years ago (see Table 3-1 below). However, these reactors use complex

engineering structures to mobilize the biomass particles and achieve high heat transfer rates. Unlike BFB

reactors which have a long operating history in oil refineries, these mechanical systems are less proven

and their ability to be scaled up may be challenging.

Another carrier-gas-free pyrolysis reactor is the “auger” or “screw” reactor which has been used for

more than 50 years in coal degassing and heavy oil coking applications(Meier et al., 2013). Due to the

relatively poor heat transfer through the auger shell, this technology is not well suited for fast pyrolysis

as it is only able to heat up the biomass relatively slowly. Auger reactors also produce lower amounts of

bio-oil as their slower heating and longer residence time characteristics favour greater solids formation


July 2014 77

(a more “charcoal-like” process). However, it has been shown that the low grade liquid and the char can

be recombined after recovery to produce pyrolysis “slurries” which can serve as an improved feed for

gasification. The energy density of the slurry can be in the range of 18-25 GJ/m3, and is typically higher

than char-free bio-oils (ca. 21 GJ/m3) or raw biomass (Dahmen et al., 2012). An example of such a

system is the Bioliq™ process from KIT in Germany which proposes to use auger derived slurries to feed

central large scale gasification facilities (Meier et al., 2013).

Although several groups around the world are pursuing biomass pyrolysis, the current production

capacity for fast pyrolysis oils is quite low. Bio-oil facilities that have been or will be operated at the

semi-commercial scale (> 50 tpd) are listed in Table 3-1 together with their characteristics and reactor

type. Most pyrolysis facilities to date are based on CFB and BFB reactor designs which, as mentioned

earlier, are relatively robust, scalable and result in relatively high yields of bio-oil.

The newly built KiOR facility in Mississippi, which is designed to operate at 500 tpd, is the first truly

commercial pyrolysis facility. This facility is distinct from other pyrolysis oil facilities as it uses catalysts

that are integrated in the CFB system (instead of sand) in order to partly deoxygenate the bio-oil prior to

upgrading to diesel and gasoline, a process commonly referred to as catalytic pyrolysis (see Section

3.3.3). This design allows integration of part of the upgrading deoxygenation into the bio-oil-making step

and is discussed in more detail in the next section which describes upgrading technologies.

A recent trend of some pyrolysis technology providers is to focus on small scale (1 -5 tpd) and mobile

pyrolizers. These units are described as mobile densification facilities that produce liquid bio-oils or

bioslurries which are intended to be subsequently transported and processed or upgraded at large,

centrally located facilities. These plants can also be co-located with oil refineries to take advantage of

co-processing opportunities. Companies that lead this trend for mobile pyrolizer systems development

include Canada’s ABRI-Tech, California’s Cool Planet and ROI in Alabama (Meier et al., 2013).


July 2014 78

Table 3-1: Commercial and pre-commercial (≥ 50 tpd) bio-oil facilities by 2014

Company Location Completion Capacity (ODT)

Application Reactor type

Fortum, Finland

Joensuu CHP plant Commissioned in Nov. 2013 (EURO 20 m own

funds and 8 m from Finnish govmnt)

100 tpd CHP fuel CFB integrated with CHP

system (Metso design)

Ensyn, Canada

Renfrew, Ontario Commissioned in 2007 100 tpd R&D resins, chemicals and CHP

fuel

CFB (Ensyn design)

Dynamotive, Canada

Guelph, Ontario Not operating 200 tpd R&D, chemicals and CHP fuel

BFB (Dynamotive

design)

Red Arrow, USA

Rhinelander,Wisconsin, USA

Commercial 2 x 50 tpd

Food flavouring/browning

products and CHP fuel

CFB (Ensyn design)

Pyrogrot, Sweden

Billerud’s pulp mill in Skarblacka

Announced end of 2012 (EURO 31 m

from NER300 = EU-ETS revenue) – expected

in 2015

720 tpd CHP fuel Not announced

Green Fuel Nordic Oy, Finland

Unspecified Q1 2014

3 x 400 tpd

CHP fuel CFB (Ensyn design)

BTG-BtL Malaysia (Netherlands)

Palm oil processing facility

(EFB feedstock)

Not operating 48 tpd Cofiring with waste Rotating Cone (BTG design)

KiOR, USA

Columbia, Mississippi, USA

Commissioned early 2013

500 tpd Drop-in diesel and gasoline

Catalytic CFB

Sources: (Bayar, 2013; BTG, 2012; Dynamotive, 2009; Ensyn, 2013a; Fortum, 2013b; Green Fuel Nordic,

2013; KiOR, 2013; Landalv, 2013; Oasmaa & Czernik, 1999; Starck, 2012)

3.2.2 Bio-oil composition

Like crude oil, pyrolysis bio-oil is a dark brown and free flowing liquid fuel (Figure 3-5), composed of

more than 300 different carbon molecules. However, chemically, pyrolysis oils are very different as they

contain about 40% oxygen compared to the typical maximum amount of 2% oxygen found in crude oil

(Speight, 2006). As discussed earlier, the oxygen content of biomass results in biofuels with undesirably

high reactivity (low chemical stability) and low energy density. For example, compared to crude oil, bio-

oil has less than 50% of this fossil fuels energy density (16-19MJ/kg vs 40 MJ/kg).

http://www.fortum.com/en/mediaroom/pages/fortum-invests-eur-20-million-to-build-the-worlds-first-industrial-scale-integrated-bio-oil-plant.aspx

http://www.redarrowusa.com/about/


July 2014 79

Figure 3-5: Pyrolysis oil sample

Source: (Empyro, 2013)

Bio-oil has a smoky odour and its chemical composition is derived from the decomposition

(depolymerisation and fragmentation reactions) of the main biomass components of lignin, cellulose

and hemicellulose (Oasmaa & Czernik, 1999). From a compositional perspective, bio-oil resembles

woody biomass much more than it resembles crude oil (Table 3-2) and thus, in some ways, it can be

thought of as “liquid wood”. In fact, “liquid wood” together with “bio-oil” are among the many

alternative names that have been used in the literature to describe pyrolysis oils (Oasmaa & Czernik,

1999). Due to this elemental resemblance to wood, bio-oil has the same effective H/C ratio of about 0.2

and would be placed at the same “step” as lignocellulosic biomass on the H/C “staircase” described

earlier. This, as will be detailed later, has implications for the hydrogen and processing needs of bio-oils

when converting them to drop-in biofuels.

When various properties of heavy fuel oil (HFO) and bio-oil are compared (Table 3-2), the differences in

the oxygen content, energy density, amount of dissolved water (up to 30%), and the poor distillation

performance (1 vs 50% residue) are quite striking.


July 2014 80

Table 3-2: Typical properties of wood pyrolysis bio-oil and of heavy fuel oil (HFO)

Physical property Bio-oil Heavy fuel oil

Water content, wt % 15-30 0.1

pH 2.5 -

Specific gravity 1.2 0.94

Elemental composition, wt. %

C 54—58 85

H 5.5—7.0 11

O 35—40 1

N 0—0.2 0.3

Ash 0—0.2 0.1

HHV, MJ/kg 16-19 40

Viscosity (at 50°C ) cP 40—100 180

Solids, wt % 0.2—1 1

Distillation residue, wt. % up to 50 1

Source: (NSF, 2011; Oasmaa & Czernik, 1999)

Water is a major component of bio-oils and its content varies depending on the initial moisture content

of the biomass and the pyrolysis conditions used. Severe pyrolysis conditions (high temperature and

residence time) remove more water but also promote vapor polymerizations and thus increase the

viscosity and solids content of the resulting bio-oil (Oasmaa & Czernik, 1999). Water is derived from

both the original water in the feedstock and from the water formed during the dehydration reactions

occurring during pyrolysis. An excessive amount of water in bio-oils is undesirable because it acts as a

heat sink during combustion and it can also promote destabilization and phase separation of the fuel

during storage. In order to minimize bio-oil’s water content, the moisture content of the biomass

feedstock is best kept below about 10 wt%. However, a certain amount of water is needed as it helps

reduce the viscosity of bio-oils. For example VTT has studied the influence of the water content on

various softwood bio-oils and found that, when the water content drops below 25%, the viscosity

increases dramatically and an unacceptable degree of bio-oil destabilization also occurs (Lehto et al.,

2013).

As well as water, the other major chemical components of bio-oil include hydroxyaldehydes,

hydroxyketones, sugars, carboxylic acids, and phenolics (Bridgwater, 2012). As was the case with water,

the amount of these compounds in the bio-oil depends on the composition of the original biomass as


July 2014 81

well as the pyrolysis conditions used. Some of these components, such as the sugars, are hydrophilic but

others such as the phenolics are more hydrophobic. Thus, most bio-oils can be considered to be micro-

emulsions. The continuous phase of the emulsion is the aqueous solution containing the polysaccharide

decomposition products and the discontinuous phase is the pyrolytic lignin (the emulsion is mainly

stabilized by weak hydrogen bonds). The breakdown of this emulsion results in the formation of two

phases, a lighter, more aqueous phase and a heavier, less aqueous (Bridgewater, 2012).

The “aging” of pyrolysis oils is measured as increased viscosity over time and it occurs through reactions

between the oxygenated carbon molecules in the bio-oil emulsion. Polymerization reactions between

double bonded components as well as esterification and etherification reactions between hydroxyl and

carbonyl groups produce high molecular weight, water-insoluble components such as gums. These

reactions lead to increased viscosity and, ultimately, to a phase separation of the bio-oil into an upper

aqueous phase (containing a higher proportion of acids and sugars) and a lower tar phase (containing

less water and a higher proportion of water insoluble solids and lignins)(Lehto et al., 2013).

The main factors that accelerate these undesirable “aging” reactions are:

Time: most bio-oils destabilize/phase-separate after storage for about 6 months or more) at

room temperature.

Temperature: very important!: For example the viscosity of a hardwood bio-oil doubled after a

year at room temperature, after a week at 60 °C and after a day at 80 °C.

Alkali char: catalyzes polymerizations thereby increasing bio-oil viscosity

(Oasmaa & Czernik, 1999)

The stability of bio-oils can be improved by removing char particles using hot filtrations (Sitzmann, 2009)

or by adding solvents (Lehto et al., 2013) to improve the stability of the emulsion. For example,

methanol blending is a relatively inexpensive upgrading method that has been shown to greatly improve

the stability of bio-oils when used as burner fuels (Diebold, 2000). In earlier work, Diebold & Czernik

(1997) showed that a 10% methanol blending reduced the bio-oil aging rate 20-fold. These and other

stabilization techniques have been discussed earlier (Diebold, 2000; Oasmaa & Czernik, 1999) but nearly

all of the options are costly and/or lead to bio-oil yield loss.

Due to the substantial amounts of non-volatile materials such as sugars and oligomeric phenolics, etc.,

found in bio-oils, they exhibit a considerably inferior distillation performance compared to petroleum


July 2014 82

heavy fuel oil (Table 3-2). The slow heating of the oils during distillation accelerates the polymerization

reactions resulting in the formation of heavy and non-volatile compounds. Heavy Fuel Oil typically

leaves 1% residue after vacuum distillation whereas a bio-oil leaves up to 50% of the starting material as

distillation residue (Table 3-2). This poor distillation performance has implications not only for further

processing of bio-oils (e.g. to drop-in fuels) but also in their use as CHP fuels.

3.2.3 Bio-oil uses

As mentioned earlier, bio-oils have great potential as fuels for burner/boilers and burner/furnaces for

stationary heat and power generation. In general, liquid fuels are easy to store, transport and combust

and they can also be pumped and fed into a burner through a spraying atomizer, thus improving heat

transfer through the fuel droplet and maximizing combustion efficiency (Lehto et al., 2013).

However, as discussed earlier, bio-oils contain large amounts of water and oxygenated compounds as

well as char particles. They also have drawbacks as combustion fuels such as low energy density, ignition

difficulties, high viscosity and instability as well as low pH and high particulate. Although bio-oils

generally produce less NOx and SOx than do fossil fuels (coal and oil), they typically emit more

particulate emissions due to the char content of the bio-oil (Lehto et al., 2013),.

Various trials have burned bio-oil in heat and power facilities (Bridgwater, 2012; Czernik & Bridgwater,

2004; Venderbosch & Prins, 2011) demonstrating the need for modified bio-oil storage and feeding

systems. The only commercial facility that regularly uses bio-oils as a burner fuel is the Red Arrow facility

in Wisconsin which uses its own bio-oils to generate heat (5 MWth) for its food flavoring production

process. Fortum’s plant (Finland) is currently (October 2013) in start-up and is targeted at bio-oil

production for use in district heating (Fortum, 2013a).

One potential use of bio-oils is to co-fire them with conventional fuels such as coal, as was

demonstrated when Red Arrow’s pyrolysis oils were co-fired in the coal power station at Manitowoc

Public Utilities in Wisconsin (USA) (Venderbosch & Prins, 2011).

As large scale internal combustion diesel engines and gas turbines are more efficient in generating

electricity than are boilers that feed steam turbines it would likely be beneficial if bio-oils could be used

in these engines in the same way as HFO is used. However, the corrosiveness and char content of bio-

oils currently limits their use in these facilities. Possible solutions to these drawbacks include the

incorporation of a low pressure fuel supply system that preheats and filters the bio-oil, a nozzle design


July 2014 83

that allows larger fuel flows and dual fuel operation, the redesign of the hot section and the use of

stainless steel parts and compatible polymeric materials. It has been noted that these modifications will

likely be costly and that there is still no guarantee that the engine will operate trouble-free when using a

bio-oil fuel (Lehto et al., 2013; Venderbosch & Prins, 2011).

Bio-oil can also be used as a fuel for gasification facilities. As gasification is more sensitive to scale than is

pyrolysis, bio-oil production could be used as a method of densification to allow the transport of

biomass over longer distances while providing the gasification facility with a better suited feedstock as

opposed to a solid fuel. The advantage of using a liquid as opposed to solid feedstock in gasification is

the same as in combustion, namely the faster heat transfer through the fuel (R. Swanson et al., 2010).

This concept of using pyrolysis oils or slurries as a gasification fuel in order to improve access to remote

biomass stocks will be reviewed in more detail in the gasification section.

3.2.4 Bio-oil standards

As mentioned earlier, the composition of bio-oils can vary significantly, depending on composition of the

feedstock as well as the pyrolysis conditions used. This influences the use of bio-oils and, in particular,

their upgrading to transportation fuels. For example, feedstocks with high extractives such as resinous

pine wood, produce bio-oils with a frothy top layer representing up to 10% of the product’s mass.

Similarly, feedstocks with a high alkali content such as grasses will likely produce bio-oils containing

alkali-rich char. Grasses are also less attractive pyrolysis feedstocks as they also have a low lignin and

high alkali ash content, both of which lead to lower bio-oil yields (about 60% compared to 75% for wood

(Bridgwater, 2012)). These examples contrast with the common perception that thermochemical biofuel

processes are “feedstock agnostic”. In practice, feedstock variation is among one of the major factors

that contribute to the heterogeneity of bio-oils and their storage stability characteristics.

To overcome the commercialization hurdles resulting from the heterogeneity of bio-oils, a set of

standards has recently been approved by ASTM. The ASTM D7544 fast pyrolysis oil burner fuel standard

was approved in 2010 for Grade G and in 2012 for Grade D bio-oils. The only difference between grades

D and G is the pyrolysis solids and inorganics (ash) content (Table 3-3). These standards qualify pyrolysis

oils as burner fuels and they provide benchmark-type minimum requirements upon which applications

and trading of bio-oils can be based.


July 2014 84

Table 3-3: ASTM standards for pyrolysis oils

Property Grade G Grade D

Gross heat of combustion, MJ/kg, min 15 15

Pour point, °C, max -9 -9

pH Report Report

Density at 20°C, kg/dm3 1.1-1.3 1.1-1.3

Kinematic viscosity at 40°C, mm2/s, max 125 125

Water content, % mass, max 30 30

Pyrolysis solids content, % mass, max 2.5 0.25

Ash content, % mass, max 0.25 0.15

Sulfur content, % mass, max 0.05 0.05

Flash point, °C, min 45 45

Source: (Nummisalo, 2012)

Further standardization efforts have been announced recently by the EU CEN. Working Group 41

(Pyrolysis oils) of CEN’s Technical Committee on: “Gaseous and liquid fuels, lubricants and related

products of petroleum, synthetic and biological origin” will, in January 2014, start developing pyrolysis

oil standards for:

a) A European Standard for a quality specification for pyrolysis oil replacing heavy fuel oil in

boilers.

b) A European Standard for a quality specification for pyrolysis oil replacing light fuel oil in

boilers.

c) A Technical Specification for a quality specification for pyrolysis oil replacing fuel oils in

stationary internal combustion engines.

d) A Technical Specification for a quality specification for pyrolysis oil suitable for gasification

feedstock for production of syngas and synthetic biofuels.

e) A Technical Specification for a quality specification for pyrolysis oil suitable for mineral oil

refinery co-processing.

The first 3 (a-c) documents above are to be given precedence and be developed as soon as possible. The

last two (d & e) are to be given lower priority and developed at a later stage or as market developments

dictate (CEN, 2014; Maniatis, 2013).


July 2014 85

3.2.5 Cost of Bio-oil

The cost of bio-oil is difficult to determine as no commercial production or trading of this product

currently occurs. Various techno-economic analyses have estimated the cost of bio-oil production with

the most recent and relevant analyses carried out by NREL (Ringer et al. 2006 and Wright et al. 2010),

KIT (Karlsruhe Institute of Technology, (Henrich et al., 2009)) and by Aston University (Bridgwater 2012).

According to Bridgwater (2012), the cost of bio-oil for a facility ($180 million in capital) processing 2000

tpd of wood ($83/odt) would be around $180/t. Assuming 18 GJ/t of bio-oil, this figure translates to a

bio-oil cost of around $10/GJ. Similar figures were estimated by Wright et al. (2010) at $10/GJ for bio-oil

from a 2000 tpd facility and an $83/t corn stover feedstock cost and by Henrich et al. (2009), $9/GJ for

bioslurry (bio-oil + char) from a 200 tpd facility and an $82/t wheat straw feedstock cost. Although

various parameters and assumptions often differ between techno-economic studies there seems to be

general agreement that bio-oil, produced at full scale facilities (2000 tpd), would cost around $10/GJ.

This is two times the assumed cost (energy basis) of the raw biomass ($82-83/odt or ca. $5/GJ) and less

than half the cost of HFO which at the time of writing was selling at about 25 USD/GJ (1000 USD/tonne).

Although these estimates are subject to uncertainties and sensitivities, they indicate that there is

currently about 50% cost margin that can be used to upgrade bio-oils to match HFO quality. A similar

percent cost margin between bio-oil and HFO was estimated by Ringer et al. (2006) who also showed

that the greatest sensitivities of the pyrolysis process were the facility size and the bio-oil yield.

Increasing the facility size from 500 tpd to 2000 tpd dropped the USD 7.62/GJ selling price down to USD

5.65/GJ while increasing the bio-oil yield from 60 to 70% dropped this price further down to USD

4.84/GJ.

Although pyrolysis has great potential as a low cost liquid fuel, it also has some disadvantages when

compared to even low grade liquid petroleum fuels such as HFO. Most of these challenges are, mainly

(directly or indirectly), related to a single overriding factor which is the relatively high oxygen content of

bio-oils. In “petroleum-like” drop-in biofuels the oxygen has to be removed and this is the primary

objective of technologies that try to upgrade bio-oils to transport fuels. Technology providers such as

UOP have claimed that, depending on the upgrading efficiency of pyrolysis oils and the price trends of

petroleum, bio-oil could become competitive in the near future. A 2005 UOP study suggested that

gasoline from bio-oil would be economically attractive if bio-oil is available at $18 per barrel ($0.11 per

liter) and crude oil sells for more than $50 per barrel ($0.31 per liter) (T. Marker, 2005). This suggests

that bio-oil would have to be produced at 63% the cost of petroleum crude on an equivalent energy


July 2014 86

basis (per GJ). This reinforces the previous point that bio-oil needs to be at least half the cost (per GJ) of

crude oil in order for the pyrolysis platform to be economically competitive.

3.2.6 Hydrothermal liquefaction

Hydrothermal liquefaction (HTL) is another thermochemical process which produces a bio-oil. However,

it is distinct from pyrolysis as it converts biomass to low oxygen bio-oil (5-20% Oxygen) and, unlike

pyrolysis and gasification, it can utilize wet biomass. The HTL process uses high pressures (e.g. 50 - 250

bar or more) and moderate temperatures (around 250-550 °C) as well as catalysts for 20-60 min to

liquefy and deoxygenate biomass (Akhtar & Amin, 2011; Elliott, 2007; Goudrian & Peferoen, 1990).

The HTL technology is not new and has been extensively studied. As early as the 1920s, Berl proposed

the concept of using hot water and alkali catalysts to produce oil out of biomass (Berl, 1944). This was

the foundation of later HTL technologies that attracted research interest especially during the 1970s oil

embargo. It was around that time that a high-pressure (hydrothermal) liquefaction process was

developed at the Pittsburgh Energy Research Center (PERC) and demonstrated (at the 100kg/h scale) at

the Albany Biomass Liquefaction Experimental Facility at Albany, Oregon, US (Elliot, 2007). At the same

time Shell Oil developed the HTU™ process in the Netherlands. The HTU™ process applied pressures in

the range 150-180 bar and temperatures in the range 300 to 350°C (Goudrian & Peferoen, 1990; Nielsen

et al., 2012). As an example, eucalyptus chips treated with HTU at 180 C and 180 bar for 6 min yielded

48.6 wt% DAF (dry and ash free basis) bio-oil, 32.8% gas and 18.6% aqueous phase. The oil contained

10% oxygen (Goudrian & Peferoen, 1990). HTL oils can be very viscous while melting points of about 80

°C have been reported (Elliott, 2007).

Due to low oil prices and other events, Shell’s HTU™ process was not commercialized. However,

technology companies such as Licella/Ignite Energy Resources (Australia), Altaca Energy (Turkey), and

Steeper Energy (Denmark, Canada) continue to explore HTL.

3.3 Upgrading Pyrolysis oils to motor fuels

As discussed previously, bio-oils can be used in various stationary heat and power applications or they

can be upgraded to drop-in biofuels such as diesel, gasoline and jet fuel grade hydrocarbons. However,

although the relatively high oxygen content of bio-oils can be tolerated for direct combustion in

stationary power applications, it is a significant problem for automobile engines and jet engines in

particular. Upgrading of bio-oils to transport fuels involves extensive deoxygenation with the major


July 2014 87

challenge being to deoxygenate the bio-oil while maintaining high conversion yields and high hydrogen

to carbon ratios in the finished fuel. Various upgrading techniques have been proposed over the last few

decades with the top two contenders being hydrotreating and zeolite cracking (Solantausta, 2011). Both

processes are catalytic and selectively promote hydrogenation reactions. Hydrotreating uses large

amounts of hydrogen to remove water from bio-oils in the form of H2O molecules. In contrast, zeolite

cracking uses no hydrogen but instead rejects oxygen in the form of CO2, thus lowering the biofuel yield.

Both technologies try to elevate the effective H/C ratio of bio-oils from about 0.2 to about 2 in order to

fit the functional properties of hydrocarbon motor fuels (see chapter 1). Virtually all of the current bio-

oil upgrading processes originated in the petroleum industry and use specialized catalysts to improve

reaction selectivity. As capital costs for upgrading bio-oils are high it would be synergistically beneficial if

existing oil refinery equipment could be used to process these biomass derived liquids. The processes

used to upgrade bio-oils resemble those used to upgrade vegetable oils to drop-in biofuels (as discussed

in Chapter 2), although pyrolysis liquids are significantly more challenging a feedstock to upgrade than

are vegetable oils (VOs).

3.3.1 Hydrotreating

Hydrotreatment is a hydrogen-intensive process for deoxygenating and upgrading bio-oils to petroleum-

like transport fuels. Earlier (in Chapter 2), it was emphasized that a lot more hydrogen is required to

hydrotreat vegetable oils (VOs) than petroleum. Even more hydrogen is required to hydrotreat bio-oils

because they contain about 40-50 wt% oxygen compared to the 10% typically found in vegetable oils. As

shown earlier on the H/C staircase diagram, the VOs have an H/Ceff ratio of about 1.8 while the bio-oil is

around 0.2 (the same as wood). Thus bio-oils require more hydrogen and processing effort to become

functionally equivalent to petroleum diesel. Similarly to hydrotreating VO’s, the targeted chemical

reaction in bio-oil hydrotreating is the rejection of oxygen in the form of H2O. This hydrodeoxygenation

(HDO) reaction of bio-oil is conceptually represented as:

C1H1.33O0.43 + 0.77H2 → CH2 + 0.43H2O (Bridgwater, 2012)

As described in the equation, the process uses about 1.5 (0.77 x 2) hydrogen atoms for every carbon

atom produced in the final fuel. As the carbon conversion is 100% (as every carbon in the bio-oil is

converted to a hydrocarbon) the resulting hydrocarbons should be of high quality since their H/C ratio is

equal to 2. However, in practice, hydrotreatment is not highly selective and the HDO reaction described

above does not take place in isolation but rather in association with other reactions which divert carbon


July 2014 88

and/or hydrogen from the targeted liquid fuel product. These reactions include polymerization and

condensation to form tars and coke, gasification to form methane or COx and reactions that form low

H/C hydrocarbons such as aromatics and olefins (Bridgwater, 2012). Thus, the low selectivity for

hydrodeoxygenation and hydrogenation reactions often leads to low fuel yields and high hydrogen

requirements.

In most of the hydrotreating processes modelled so far the biomass to fuel yield is around 25% mass

(55% energy) when hydrogen is provided externally and 15% (33%) when hydrogen is produced by

gasifying the biomass (Brown, 2011; Bridgwater, 2012; Dynamotive, 2013). . However, as will be

described briefly, these relatively poor carbon yields and hydrogen use efficiencies can be improved

through the development of more selective catalysts and optimized processes.

In Section 1.10, it was shown that, in the US, a doubling of refinery hydrotreating capacity would be

needed in order to meet the requirements for processing crude oils of deteriorating quality. A similar

doubling or even tripling of today’s US refinery hydrogen generation capacity would be needed if the US

refineries were to provide the hydrogen amounts needed to meet the US RFS cellulosic advanced

biofuels mandate exclusively with pyrolysis-derived diesel. The US RFS has mandated 15 billion gallons

per year of cellulosic advanced biofuels by 2022 (includes pyrolysis diesel). If this entire mandate was to

be met by biomass pyrolysis-derived fuels, around 250 facilities producing 60 million gallons per year

(MGPY) of diesel/gasoline blendstocks would be required. According to the latest techno-economic

analysis by Jones at al. (2013) on biomass fast pyrolysis to automotive fuels, each of the 60 MGPY

pyrolysis drop-in biofuel facilities would require a hydrogen generation capacity of 44 million standard

cubic feet (SCF) per day. Assuming this hydrogen requirement would come from existing US refineries, a

capacity of 11 billion SCF would be required by 2022. Current US refinery hydrogen generating capacity

is only ca. 3 billion SCF per day (US EIA, 2013b).

As described earlier, owing to the highly heterogeneous, oxygenated and reactive nature of bio-oils,

their hydrotreatment is a lot more complex than that of petroleum. In oil refineries, hydrotreatment is

mainly used to remove sulfur from petroleum feeds in a process known as hydrodesulfurization (HDS).

The process conditions include temperatures that range between 300 and 600 °C, hydrogen pressures of

35 to 170 bar and liquid hourly space velocities (LHSV) of 0.2 to 10 per hour. The catalysts used in

petroleum HDS are typically sulfided Co-Mo and Ni-Mo supported on porous alumina or aluminosilicate


July 2014 89

matrices. Unfortunately, as described below, these conditions are not suitable for processing bio-oils for

a number of reasons:

Sulfided Co-Mo and Ni-Mo catalysts, when in contact with bio-oils, are rapidly stripped of their

sulfur and require constant resulfurization (addition of H2S) to prevent catalyst deactivation

(Huber, 2007). This deprives bio-oils of their low sulfur content advantage (Wang et al., 2013)

Alumina supports create an acidic environment and they are not stable in the presence of water

(irreversible dealumination) (Mortensen et al., 2011)

Bio-oils are unstable at high temperatures as they can rapidly become viscous and eventually

phase separate.

Bio-oils tend to form coke residues, particularly in acidic environments and at high temperature

and pressure. Coke is undesirable as it deactivates the catalysts by depositing on their active

sites and it can severely plug reactor components (Wang et al., 2013).

The water in bio-oil inhibits hydrotreating by modifying and deactivating the catalysts and by

adsorbing onto active sites (Furimsky & Massoth, 1999). Aside from the water content of bio-oils

(up to 30%), more water is produced upon hydrotreatment.

These and other limitations have motivated the search for hydrotreating processes and catalysts that

are better suited to the highly oxygenated and heterogeneous nature of bio-oils. Early research focused

on adjusting process conditions and working with model bio-oil mixtures while using the same sulfided

catalysts that oil refineries use for desulfurization (Corma et al., 2007). Although these alumina

supported Co-Mo and Ni-Mo catalysts have various problems in processing bio-oils, they improve

hydrotreating selectivity and they are widely available at relatively low cost.

As mentioned previously, bio-oils are thermally unstable and they have to first be pretreated at lower

temperatures in order to form a stable oil intermediate that can then be further hydrotreated at high

temperatures. Although single stage hydrotreating of bio-oil at high temperatures has been attempted,

it resulted in a heavy, tar like product (de Miguel Mercader et al., 2010; Jones et al., 2009) due to the

polymerization, charring and eventually coking reactions which, at high temperatures, take place faster

than the desired hydrotreating reactions. However, to achieve effective hydrotreatment high

temperatures and hydrogen pressures at extended reaction times (up to 4 hours) (Elliott, 2007) are

often required. To fulfill these disparate condition requirements for stabilization and complete


July 2014 90

hydrotreatment of bio-oils, a two-stage bio-oil upgrading approach is commonly used (Elliott, 2007;

Jones et al. 2009). The first, mild, catalysed hydrotreatment stabilizes the bio-oil and a second, higher

severity hydrotreatment stage, deoxygenates the fuel to transport-grade liquids. The first

hydrotreatment typically forms at least two phases, one hydrophobic and one hydrophilic and

effectively separates out a large proportion of the water within the bio-oil. The resulting hydrophobic

liquid is more stable and amenable to further catalytic upgrading.

When earlier workers (Centeno et al., 1995; Ferrari et al., 2001) investigated the fundamentals of bio-oil

hydrotreatment using traditional sulfide molybdenum catalysts on model bio-oil compounds such as

ketones, esters and phenolics (while alcohols and carboxylic acids where formed in the process), they

concluded that ketones react first at lower temperatures (> 200 °C) to form alkenes while carboxylic and

phenolic groups are converted at higher temperatures (> 300 °C). This early work lead to a proposed

reactivity scale for the major components of bio-oils; the scale is plotted in Figure 3-6. The olefins and

other double bond species are the most reactive and can be hydrogenated to more stable components

such as alcohols and alkanes at temperatures around 250 °C and below. Alcohols are dehydrated to

olefins at temperatures closer to 300 °C while carboxylic groups are more recalcitrant than alcohols and

aromatics are the most recalcitrant and will only react at temperatures in the vicinity of 400 °C. This

reactivity scale is a very useful rule of thumb in the absence of precise reaction kinetics.

Figure 3-6: Reactivity scale of oxygenated compounds under hydrotreatment conditions. Source: copied from Wang et al., 2013, based on work from Delmon and co-workers (Centeno et al., 1995; Ferrari et al., 2001) as adapted and plotted by Elliott, 2007.


July 2014 91

As noted earlier, coking is a major problem during hydrotreatment as it can lead to catalyst deactivation

and reactor plugging. In general, the parameters that promote coking are high temperatures, low

hydrogen pressures, high acidity, and the presence of low H/C components such as phenolics, alkenes

and highly oxygenated carbon molecules (Huber, 2007; Mortensen, 2011). Double bond molecules such

as olefins, ketones and aldehydes are particularly prone to polymerization and coking. Fortuitously,

these species can be hydrogenated relatively easily during the first, low severity, hydrotreatment stage.

This improves the thermal stability of the resulting bio-oil before the second hydrotreatment step.

Refinery HDS catalysts promote the formation of coke by creating an acidic environment and promoting

the formation of aromatics. Aromatics are desirable up to certain concentrations since they form part of

transport fuel blends, particularly gasoline (40% aromatics) (Bauen, 2009). However, aromatics are a low

H/C ratio species and can act as precursors for coking reactions upon upgrading. The hydrogenation of

aromatic rings is the most challenging as it requires high temperatures and hydrogen pressures (around

4.0 to 8.0 MPa of H2) as well as highly active catalysts such as precious metals (Wang et al., 2013).

Another way in which hydrogen can reduce coke formation is by converting catalyst-absorbed reactive

species, such as alkenes, to stable molecules such as alkanes. In general the presence of hydrogen

appears to play a pivotal role in minimizing the formation of coke.

While this more fundamental work has built an invaluable body of knowledge around the

hydrotreatment and coking mechanisms of bio-oils, these studies have been mainly based on “model”

as opposed to “real” bio-oils (Butler et al., 2011). In contrast, Elliott and co-workers (Elliott, 2007) used

sulfide Co-Mo and Ni-Mo catalysts on real bio-oil substrates and developed a two stage process that

brings the oxygen content of the bio-oil down to < 1 wt%. The first stage of the process is conducted at

270 °C and 136 bar and the second at 400 °C and 136 bar. Consistent with the known challenges of these

catalysts, the authors reported catalyst deactivation and formation of gums as major drawbacks of the

process.

More recent research has focused on developing catalysts that may circumvent the challenges

encountered with traditional HDS catalysts such as CoMo and NiMo supported on alumina materials.

Precious metals such as Ruthenium, Palladium and Platinum have been assessed as bio-oil

hydrotreatment catalysts (Bridgwater, 2012). These metals performed better than CoMo and NiMo

catalysts in terms of both hydrocarbon yields and H/C ratio of final product (Lin et al., 2011; Wang et al.,


July 2014 92

2013; J. Wildschut et al., 2009). They are also more stable, less acidic and do not promote coking

particularly when supported on non-acidic carbon. The company, UOP, has been a leader in using

precious metal catalysts for hydrotreatment of petroleum. Together with PNNL they have assessed the

potential of non-sulfided metal catalysts such as Ruthenium on bio-oils. Ruthenium seems to be the

lowest cost and most promising of the precious metal catalysts assessed so far (J. Wildschut et al., 2009;

J. Wildschut, Melian-Cabrera et al., 2010; J. Wildschut, Iqbal et al., 2010). When Lin et al. (2011) assessed

various precious metal catalysts using the model compound guaiacol, they reported that the Rh-based

catalyst showed the best HDO activity and a preference to saturate benzene rings. Although Ruthenium

is less expensive than Palladium and Platinum (similar price to Gold!) on June 28, 2013 the spot price for

Ru was about USD $3 million/t which is more than a 100 times the same day price of Cobalt (ca. USD

$30,000 USD/t), Nickel (ca. USD $10,000/t) or Molybdenum (ca. USD $20,000/t) (IndexMundi, 2013;

InvestmentMine, 2013). Although precious metal catalysts are more favoured for bio-oil

hydroprocessing, as they are more active in comparison to NiMo and CoMo based catalysts, their cost is

so prohibitive that their use in industrial applications may be very limited.

The ability to recycle and the stability of Ru/C catalysts has been challenged by Wildschut (2009). When

he conducted three successive hydrotreatment reactions (200 bar, 350 oC and 4.3 h each) where the

catalyst was reused after repeated acetone washes, he found that the activity of the catalyst

deteriorated even after the first repeat. After 2 repeats the oil yields dropped significantly (55 to 30%-

wt.), whereas the amount of gas phase (5 to 11%-wt.) and solids (3 to 20%-wt.) increased, all indicating

significant catalyst deactivation. This deactivation mostly affected the ability of the process to

hydrogenate while it did not affect much of its ability to deoxygenate. The deactivation was attributed

to sintering and coke formation on the surface of the catalyst. It would therefore be desirable if the

catalysts could have been regenerated with a more effective technique than acetone washing.

The prohibitive price of precious metals means that novel catalysts have to be designed which will

achieve high hydrotreating activity at lower cost. Although phosphide catalysts have been suggested as

alternatives to sulfide catalysts they face similar issues. Once in contact with water they form

phosphates which can deactivate the active sites on the catalyst (Wang et al. 2013).

Other than deoxygenation, the hydrogen treatment of bio-oils has many favourable side effects such as

decreasing its water content, increasing its energy density (from 18 MJ/kg in crude bio-oil to 40 MJ/kg in

hydrotreated bio-oil), decreasing its bulk density (from >1 in bio-oils to <0.8 in deoxygenated bio-oils),


July 2014 93

decreasing the coking propensity, and decreasing its viscosity (from > 100cps in raw bio-oil to <5 cps in

bio-oil that contains <5% oxygen) (Elliott, 2007). All of these advantages result in higher yields and the

higher H/Ceff ratio of the final transport fuel product. These benefits are generally absent in any bio-oil

upgrading technologies that do not entail any hydrogen inputs.

3.3.2 Zeolite cracking

Zeolite cracking is a process used in the Fluid Catalytic Crackers (FCCs) of oil refineries and it has

potential in bio-oil upgrading as a non-hydrogen consuming, non-pressurized alternative to

hydrotreatment. The main deoxygenation mechanism of zeolite catalytic cracking is the rejection of

oxygen in the form of coke and CO2. The conceptual reaction of this mechanism is summarised below:

C1H1.33O0.43 + 0.26O2 → 0.65CH1.2 + 0.34CO2 + 0.27 H2O

When this formula is compared with the earlier equivalent formula for hydrotreating it is apparent that,

in the absence of hydrogen, as occurs in zeolite cracking, bio-oil upgrading is poor. The theoretical

carbon yields for zeolite cracking are low (65% compared to 100% in hydrotreatment) and the

hydrocarbons produced have a low H/C ratio (1.2 compared to 2 for hydrotreatment). This low H/C ratio

indicates that the bio-oil is rich in aromatics and olefins and that the resulting fuel will have a low

heating value, typically about 20-25% lower than crude oil (Balat et al., 2009; Mortensen et al., 2011).

Similar to what occurs in hydrotreatment, the cracking reaction takes place alongside other undesirable

reactions such as polymerization and coking which results in the diversion of some of the carbon from

the targeted liquid biofuel. Thus, even in the presence of catalysts, cracking typically results in bio-oil-to-

fuel yields in the range of 14-23 wt% of bio-oil (Balat et al., 2009), which is much lower than the

theoretical 45 wt% yield which can be calculated from the equation above. This is largely because 26-39

wt% of the starting bio-oil goes towards the formation of solid tar and cokes (Balat et al., 2009).

Zeolites such as ZSM-5 and HZSM-5 are made of a highly porous aluminosilicate matrix and, as a result,

they are typically not stable in the presence of bio-oils at high temperatures and pressures. Zeolite

catalysts such as ZSM-5 have a strong acidity, high activities and shape selectivities which work well for

upgrading petroleum feeds. However, for bio-oils, zeolite cracking poses severe catalyst coking and

deactivation issues.

On the more positive side zeolite cracking requires no hydrogen gas and can operate at atmospheric

pressures. This means that FCC-type systems can be used for bio-oil processing where the heavily coked


July 2014 94

catalyst can be rapidly regenerated in the FCC combustor. These systems have great potential to utilise

the coke formed on catalysts as a fuel for heat and power generation. However, these systems often

convert more biomass carbon to thermal energy than to liquid fuel products. Operating costs of FCCs

are higher than regular fixed bed reactors because the faster recycling of carrier gas needed to

regenerate the rapidly coked catalyst is highly energy intensive.

Other workers have recently (Vispute et al., 2010) proposed an approach that involves a mild

hydrotreating step prior to zeolite cracking. The advantage of this approach is that it converts the most

reactive oxygenated compounds, the carbonyls, to more thermally stable alcohols. In zeolite cracking

carbonyl functionalities go directly to coke formation whereas alcohols contribute to the formation of

valuable molecules such as olefins and aromatics. The introduction of a mild hydrotreatment step prior

to zeolite cracking appears to result in bio-oil conversion yields (aromatics) as much as three times

higher than direct zeolite cracking. This is particularly desirable for BTX (benzene, toluene, xylene)

production which is the target of the Huber group’s spin off company AnelloTech (vide supra). However,

the alkane yields are low and, while this technology may be relevant to making aromatic fractions for

gasoline, it is not directly applicable to the production of the longer chain hydrocarbons that are needed

for the production of diesel and jet fuels.

Cracking of bio-oils has also been attempted using platinum catalysts. Fisk et al. (2009) achieved a low

oxygen bio-oil (2.8%) which was predominantly composed of aromatic and very few non aromatic

oxygenates. The oil yields of the process were rather low (10-20% wt of bio-oil was converted to stable

oil) when compared to hydrotreatment upgrading (ca. 40%) (Butler et al., 2011; Fisk et al., 2009).

3.3.3 Catalytic pyrolysis

Catalytic pyrolysis is a process that combines pyrolysis with zeolite cracking in a single step (Butler et al.,

2011). Due to the challenges of catalyst coking, the preferred reactors for these systems are again the

FCC-type reactors. The main goal is to use the catalysts inside the pyrolysis reactor to increase the

reaction selectivity towards deoxygenation, thus producing a less oxygenated bio-oil. As occurs in zeolite

cracking, the oxygen is primarily removed by decarboxylation, but at the expense of bio-oil yield and

diminished bio-oil properties. Compared to non-catalyzed pyrolysis, catalytic pyrolysis typically yields a

more viscous bio-oil that represents only 30-40% of the starting biomass (60-75% for non-catalyzed

pyrolysis). A recent DOE funded project at RTI International showed that catalytic pyrolysis (vapor phase

upgrading) yields dropped to as low as 6-13% compared to 48% in uncatalyzed pyrolysis (DOE, 2011).


July 2014 95

Again the loss in bio-oil yield is compensated by an increase in power generation through combusting

the excess coke and gas that is formed at the expense of liquid yields.

Despite the technical difficulties and low liquid yields of catalytic pyrolysis, the process is commercially

attractive as it holds promise for capital cost savings and the production of bio-oils that are more

deoxygenated and chemically stable as well as generating large amounts of renewable power. As

described below, catalytic pyrolysis forms a key part of the processes proposed by several of the leading

biofuel companies:

Annellotech, a spin-off company from Professor George Huber’s group at the University of

Massachusetts, is licensing a fluidized bed catalytic pyrolysis platform which produces a bio-oil

rich in benzene rings (hence “Anello” Tech which in Latin means “ring”). The aromatic character

of this biomass derived liquid means it is good feedstock for the subsequent production of BTX

(Benzene, Toluene, Xylenes). The company is primarily focussed on these value-add chemicals

(Anellotech, 2013).

The Gas Technology Institute in Chicago, USA, has recently developed a catalytic pyrolysis

known as catalytic hydropyrolysis (or integrated hydropyrolysis and hydroconversion) and

marketed as “IH2”. In this process biomass is converted in a fluidized bed of catalyst under

hydrogen pressure of 20–35 bar and temperatures of 350–480°C and in the presence of

catalysts (T. L. Marker et al., 2012). Although this process produces bio-oils with a low oxygen

content (e.g. <3 wt%, T. L. Marker et al., 2012) in a single reaction step, scaling up and high

hydrogen consumption remain as potential challenges of hydropyrolysis.

KiOR uses catalytic pyrolysis (FCC-type reactor and proprietary catalysts) followed by

hydrotreatment to produce transport fuels. As mentioned earlier, KiOR has recently built the

world’s first commercial pyrolysis drop-in biofuel plant and the company has also claimed that

the catalytic pyrolysis bio-oil can be directly inserted into a petroleum refinery for further

processing. KiOR’s proprietary biomass fluid catalytic cracking (BFCC) process should work in

both standalone pyrolysis facilities and in refinery co-processing. Although little is publically

known about the technical details of the BFCC process it has been reported that clay materials

and pre-impregnation of biomass with nanocatalysts are involved (Bridgwater, 2012).


July 2014 96

3.4 Refinery integration of pyrolysis platforms

As mentioned earlier, the majority of the processes and catalysts used to upgrade pyrolysis oils originate

in the oil refining industry. It has also been suggested that pyrolysis oils or their derivatives could be

“dropped into” existing refineries for final processing (Corma et al., 2007; Solantausta., 2011). The main

benefit of this approach is capital cost savings by utilizing facilities and off-take infrastructure that has

already been built. For example, the USDA “Regional Roadmap to Meeting the Biofuels goals of the

Renewable Fuels Standard” (2010) concluded that 527 new biorefineries would be needed to meet the

requirements of the RFS 2, at a cost of about 168 billion USD (Weyen, 2012). A big part of this capital

cost could be avoided if biomass intermediates could be upgraded to biofuels using existing oil refinery

equipment. It should be noted that, in the US, refinery utilization is expected to decrease as the US EIA

predicts a reduction in refinery throughput over the next decade (EIA, 2009). Another important trend

that is projected for the next three decades is that refineries in the US and around the world will be

producing less gasoline and more diesel and jet fuels (more middle distillates). This shift translates to

refineries directing petroleum feed away from FCC units and towards hydrocracking units. According to

the US EIA 2013 Annual Energy Outlook (EIA, 2013), the already decreased utilization of FCCs (83% in

use in 2011) in US refineries is expected to decline further and approach 62% in 2040. In contrast the US

hydrocracking capacity is expected to increase from 1.8 million bpd in 2012 to 3 million bpd in 2040.

It has been suggested that oil refiners could either dedicate whole process units such as hydrotreaters

exclusively to bio-oil processing or they could co-process bio-oils together with petroleum feeds (Corma

et al, 2007, Egeberg et al., 2010). Dedicating whole refinery units to upgrading bio-oil derivatives would

save capital costs and avoid complications of co-processing. However, candidate refinery units for

biomass liquids processing such as hydrotreating and hydrocracking facilities are very large scale and as

noted earlier, they typically process around 100,000 barrels of fuel per day (see Chapter 1 for details on

these refinery processes). Commercial pyrolysis facilities are usually envisioned to be about 30 times

smaller at around 3,000 barrels per day, at a scale large enough to benefit from economies of scale

while small enough to avoid transporting bulky and wet biomass over prohibitively long distances

(Stephen et al., 2010). Thus sourcing, transporting and utilising the biomass feedstock needed to occupy

a whole refinery unit will be challenging. Thus, co-processing is a more likely integration pathway than

dedicating entire refinery units to biomass feeds. Co-processing also has the advantage that small

amounts of biomass derived liquids can be blended with petroleum feeds in order to mitigate the

problems that come with neat pyrolysis oil processing.


July 2014 97

As mentioned earlier, neat pyrolysis oils cannot be readily co-processed with petroleum feeds as they

typically contain up to 30% water and 40% oxygen and are thus not miscible with the apolar petroleum

liquids (Venderbosch & Prins, 2011). As the oxygen content of bio-oils also increases coking and

deactivation of zeolite and HDO catalysts they cannot be readily inserted in oil refineries before at least

partial deoxygenation (hydrotreated). However, it will be important to deoxygenate only enough to

meet the minimum requirements of the refinery since deoxygenation gets disproportionately costlier

when approaching oxygen-free bio-oils (Elliott, 2007; Ringer et al., 2006).

Once the oxygen content of the bio-oil has been reduced by hydrotreatment, it becomes a liquid

hydrocarbon intermediate (such as hydrodeoxygenated oil, (HDO)) that can potentially be inserted into

an oil refinery. As HDO bio-oils, even when partially deoxygenated, are unstable at 400 °C or 500 °C

(temperatures that are often used in petroleum distillation) they cannot be directly inserted with crude

oil at an early process stage of the refinery. Thus bio-oil insertion is likely to occur at the refinery’s

hydroprocessing (hydrotreatment and hydrocracking) or fluid catalytic cracking reactors. As described

earlier, these two processes are similar to the processes used for hydroprocessing and zeolite cracking

of neat pyrolysis oils in stand-alone setups. A simplified schematic showing HDO bio-oil insertion points

(red arrows) within a typical refinery is outlined in Figure 3-7.

Figure 3-7: Refinery insertion points (red arrows) for HDO Bio-oils

Source: adapted from (US EIA, 2007)

Fluid catalytic cracking

Crude oil

Distillatio

n

to

we

r

Vacuum unit

Light ends

Heavy ends

Reformer Gasoline

Jet

Hydrotreatment Diesel, Jet

Hydrocracker Diesel, Jet

Gasoline

Coker Coke

Hydrotreatment Gasoline

Diesel, Jet

Pro

du

ct blen

din

g

DISTILLATION (CATALYTIC) UPGRADING BLENDING

HDO Bio-oil

Naphtha

Gasoline


July 2014 98

It is important to clarify that the use of the term hydrodeoxygenated oil (HDO) is not well defined as it

simply refers to a bio-oil that has been stabilized through hydrotreatment. However, the degree of

hydrotreatment can vary markedly and it largely depends on the co-processing insertion point and the

blending ratio. The FCC insertion point can take more oxygen while hydrocrackers are far more sensitive

to oxygen as they operate under very high temperatures and pressures. These two HDO-petroleum co-

processing strategies are discussed in detail below.

3.4.1 Fluid Catalytic Cracking (FCC) co-processing

Fluid catalytic cracking (FCC) of bio-oil mixtures with petroleum feeds resembles the earlier described

zeolite cracking process. It requires no hydrogen and it is more tolerant than hydroprocessing to higher

oxygen content biomass liquids (Solantausta, 2011). Unlike hydroprocessing reactors, FCCs operate at

atmospheric pressure and they can efficiently regenerate the coke deposited on the catalysts by

circulating them through their fluidized bed combustor loop. As described earlier, the combustion of this

excess coke generates heat and thus benefits the energy balance of the refinery process. Various groups

have carried out FCC-type trials of co-processing bio-oil (or bio-oil model mixtures) in blends with

petroleum vacuum gas oil (VGO) (Bezergianni et al., 2009; Huber & Corma, 2007; Lappas et al., 2009).

This work showed that, when partly deoxygenated pyrolysis oils were blended with VGOs, lower H/C

ratio products were produced as compared to VGO processed alone. It was apparent that heavier (coke

and tar) and lighter (gasoline and gases) fractions were produced at the expense of middle distillates

while the gasoline produced was generally poorer in saturates and richer in aromatics.

Encouraging results were obtained after the FCC co-processing of HDO bio-oils with heavy petroleum

fractions. Fogassy et al. (2010), successfully mixed 20% of an HDO (which contained 20% oxygen) with

petroleum VGO and then co-processed the mixture in an FCC arrangement. Although the gasoline yields

were comparable to those obtained with pure VGO FCC processing, overall the H/C ratio of the product

was low and the amount of coke, aromatics and olefins produced were significantly higher. Even more

encouraging results were obtained by de Miguel Mercader et al. (2010) who successfully co-processed

an HDO (28% oxygen) with a Long Residue heavy petroleum feed supplied by Shell Inc. The product

distribution was similar to the Long Residue processed alone with a less-than-expected increase in coke

and dry gas production compared to the base feed. However, the H/C of the final product was

significantly lower. These workers concluded that the effective H/C ratio of the HDO/petroleum blend is

a good indicator of the quality of the resulting FCC product (de Miguel Mercader et al. 2010). Both of

these recent studies (de Miguel Mercader et al. 2010, Fogassy et al. 2010) reported a synergistic


July 2014 99

beneficial effect of blending HDOs with petroleum feeds since a more than proportional decrease in

coking propensity of HDOs was observed when they were blended as compared to when they were

processed pure. The synergistic effects were mostly attributed to the transfer of hydrogen from the

petroleum to the biomass feed during co-processing. It was apparent that this process hydrogenated

coke-inducing oxygenated compounds at the expense of the overall H/C ratio of the blend. When

industrial trials were subsequently performed to test if these highly oxygenated pyrolysis HDOs could be

co-processed at a large scale, the 20% HDO blending proved to be challenging while, although the 5%

HDO blend were found to be technically feasible, increased coking was observed (Solantausta, 2011).

3.4.2 Co-hydrotreating

Co-hydrotreating of HDOs is another co-processing strategy that could be used to insert HDO bio-oils

into oil refineries. Although there is limited experimental data on co-feeding of real bio-oils with

petroleum feeds in hydrotreating units studies when using model compounds, the co-hydrotreating of

HDO bio-oils resulted in similar problems to those observed when co-hydrotreating vegetable oils. These

included, competition with hydrodesulfurization, increased coking and catalyst deactivation as well as

increased hydrogen demand and potential irregular hydrogen pressure drops inside the reactor (Butler

et al. 2011). Other work also reported (Bui et al. 2009) that, when gas oil was co-hydrotreated with

guaiacol over a CoMo catalyst, the gas oil did not get fully hydrotreated until the guaiacol was

completely hydrodeoxygenated. This indicated that the deoxygenation reactions were prioritized at the

expense of desulfurization reactions which are a highly undesirable outcome from a refiner’s

perspective. In today’s oil industry where sulfur regulations are becoming increasingly rigorous, a

compromised or slow desulfurization unit is problematic. As mentioned earlier, hydrotreatment units

are rather sensitive to oxygen and unlikely to process bio-oils with an oxygen content that exceeds

about 5% at blending ratios of less than 10%. These limitations could be overcome if improved catalysts

such as the recently developed NiCu catalyst and other non-disclosed catalysts patented recently by

participants of the BIOCOUP project in Europe (Solantausta, 2011) can be commercialised.

As also mentioned earlier, hydrocracking is a more severe form of hydrotreatment and it aims at

cracking the heavy portions of bioderived hydrocarbons. This process follows hydrotreating in an oil

refinery and it is even less tolerant to oxygen than hydrotreatment (due to higher pressures and

temperatures). Hydrocracking units can be leveraged by biomass derived HDO oils that have been

extensively deoxygenated (<2%) and some recent techno-economic studies describe design cases where


July 2014 100

highly hydroprocessed HDOs (Jones et al., 2009) are finished in the hydrocracker unit of an oil refinery in

order to save capital equipment costs.

Much of the recent advances in better understanding the co-processing of HDOs with bio-oils in oil

refineries have been a result of the work carried out under the European 6th Framework program

BIOCOUP. This 5 year program ended in 2011 and a summary of important conclusions is listed below

(Solantausta, 2011):

Despite their low oxygen content, bio-oils catalytically upgraded without hydrogen

(Decarboxylated oils or DCOs) or upgraded without catalyst or hydrogen (High pressure thermal

treatment oils HPTTs) could not be co-processed while HDO bio-oils could.

Despite their relatively high remaining oxygen content (between ~10 and 30 wt.%) and low

hydrogen use (only ~100-300 NL/kg), all HDO oils could be co-processed with good results in a

laboratory scale fluidised bed reactor (SP3 project). Considering that deoxygenation was initially

thought to be the main goal of upgrading this was a surprising result.

Important criteria to allow successful co-processing of the HDO oil were, low coking tendency

(measured as MCRT), a high H/C and H/Ceff ratios and a not-too-high molecular weight of the

HDO oil.

Although some of bio-oil stabilization was accompanied by a limited deoxygenation, substantial

deoxygenation in itself (requiring ~ 800 NL hydrogen/kg pyrolysis oil) was not needed.

Based on these conclusions the BIOCOUP project suggested an integrated strategy of bio-oil co-

processing using the FCC as the refinery insertion point. The strategy entails using the refinery’s FCC to

introduce partially hydrotreated HDOs (about 20% oxygen remaining after using about 200 L of H2 per L

of oil upgraded) and prioritizing emissions and cost savings. Reduced hydrogen usage is achieved by

prioritizing the hydrotreatment of the most unstable (rather than all) bio-oil components. Capital cost

savings would then be achieved by using the existing FCC infrastructure early in the process flow.

Although the yields might be lower, the LCA and techno-economic analyses demonstrated that the FCC

based BIOCOUP approach compared favourably with other, alternative approaches in terms of both

GHG emissions and economic returns. Specifically it was demonstrated that, using the BIOCOUP strategy

on $7/GJ wood, an $18/GJ diesel could be produced ($25/GJ is the equivalent for the August 2013 diesel

spot price) (IndexMundi, 2013; Solantausta, 2011).


July 2014 101

It is apparent that both FCCs and hydroprocessing refinery units can accept bio-oils that have been

partially deoxygenated (HDOs). However, the two facilities differ in their relative suitability for biofeed

insertion (Table 3-4). It has been shown that FCCs can handle lower grade (up to 20% oxygen) HDOs

without the need for hydrogen and this results in low conversion yields, large amounts of waste energy

production and lower value end products (low H/C ratio) which will contribute mostly to marine and

bunker fuel blendstocks. In this regard FCC can be viewed as the “workhorse” of bio-oil co-processing.

Alternatively, hydroprocessing could be considered the refinery’s “boutique” upgrading unit as it

requires more deoxygenated bio-oil co-feed (max of 3-5%) and it is designed to produce higher grade

diesel and jet fuels.

Table 3-4: Comparison of FCC and hydroprocessing as refinery co-processing platforms for bio-oils

FCC (Fluid Catalytic Cracking) Hydroprocessing

(hydrotreating followed by

hydrocracking)

Tolerable oxygen content in bio-oil feed Up to 20% (<5% blend) No more than 3-5%

Pre-refinery hydrotreatment Single stage/mild Two-stage/severe

Hydrogen use (NL / L of bio-oil) ca. 300 >800

Yields Low (10-15% in standalone) High (25% in standalone)

Downstream contamination risk with oxygenates

High Low

Opportunity for power generation through coke combustion

High Low

Chemistry of product Favors aromatics (low H/Ceff

ratio)

Favors short paraffins (high

H/Ceff ratio)

Fuel pool (most contribution) Gasoline and Marine fuels Middle distillates: Diesel

and Jet fuels

Source: (adapted from Solantausta, 2011)

3.5 Techno-economics and sensitivities of upgrading bio-oils

Although the technology for upgrading pyrolysis oils to biofuels is still evolving, some estimates of the

overall costs involved are provided in four recent techno-economic analyses (Jones et al., 2009; Wright

et al., 2010, Brown et al., 2012, Jones et al. 2013). While direct comparisons between techno-economic

analyses are often problematic, these four studies have some common assumptions such as 2000 tpd


July 2014 102

processing capacity and a 10% ROI (cost year base is 2007 for the first 2 studies and 2011 for the two

more recent studies, but this difference is not expected to significantly affect the gross comparisons as

the Chemical Engineering Plant Cost Index for equipment increased by 13.6% between December 2007

and April 2011 (Brown et al, 2012)). The 2009 study by the US DoE Pacific Northwest National

Laboratory (PNNL) (Jones et al., 2009) was based on poplar bio-oil upgraded with a 2-stage

hydrotreatment system using a conventional CoMo HDS catalyst. This study, which could be considered

the most detailed publically available to date, estimated a pyrolysis diesel minimum fuel selling price

(PFSP) of $2.04/gal ($0.54/L) with this price dropping to $1.74/gal ($0.46/L) when hydrogen is imported

to the facility rather than produced on site from biomass. Importing hydrogen (made from natural gas)

was considered more favoured in the near term since the conversion of biomass to hydrogen is

currently inefficient and costly. There are also considerable capital cost savings achieved when hydrogen

is imported as opposed to making it from biomass as the $303 million total project investment (TPI)

drops to $188 m when hydrogen is imported from an external source. The TPI breakdown described in

Table 3-5 shows how, out of the total $303 million cost of the plant, only $92 million is the cost of the

fast pyrolysis process components while the remaining costs are mostly related to the hydroprocessing

infrastructure. Thus very substantial capital costs could be saved if the upgrading operations could make

use of the hydroprocessing and hydrogen generation facilities in an existing oil refinery. The PNNL study

takes a somewhat conservative (and appropriate) approach and uses the refinery only as a source of low

cost hydrogen and to perform the final hydrocracking of the heavier portions of the highly

deoxygenated (2% O2) HDO. This approach results in a capital savings of $115 m.

Some of the assumptions used in the 2009 PNNL study are worthy of discussion, such as feedstock costs

of $51/t for a 2000 tpd facility. More recent studies have suggested that costs of around $80/t are more

likely considering the logistic challenges and the competing end uses for the biomass. The PNNL

sensitivity analysis indicated that an $80/t feedstock cost would add about $0.20/gal ($0.05/L) to the

MFSP. Possibly the biggest concern regarding the PNNL analysis is the assumption that the hydrotreating

catalyst will last at least 1 year. This is about the average life expectancy of catalysts used in refineries

that hydrodesulfurize traditional petroleum cuts. As described earlier, catalysts that process bio-oils

have so far proven to have much shorter life expectancies due to extensive coking. To date, no study has

demonstrated more than 200 h (8 ½ days) of continuous catalyst operation when hydroprocessing bio-

oils (de Miguel Mercader et al., 2011). The sensitivity analysis of the PNNL study indicated that halving

the catalyst lifespan to 6 months added $0.15/gal ($0.04/L) to the base MFSP (minimum fuel selling


July 2014 103

price). However, a catalyst lifespan in line with current technology would be about 20 times shorter (8 ½

days vs 6 months).

The 2009 PNNL study was recently (November 2013) updated and some of the main cost estimates and

assumptions are listed in Table 3.5. The new analysis (Jones et al., 2013) was based on a similar process

sequence and conversion yields reported in 2009 and the hydrotreating catalyst lifetimes are again

assumed to be 1 year. The 2013 study makes a clear statement that it is not based on available

technology but on technology that is expected to be available by the year 2017 (Jones et al., 2013). This

new study shows significantly higher capital and operating costs for producing diesel and gasoline via

fast pyrolysis of biomass. As shown in Table 3.5, the $700 million TPI and $3.39/gal ($0.90/L) MFSP

reported in 2013 compares to a $303 million and $2.04/gal ($0.54/L) MFSP reported in 2009 (own

hydrogen scenario) for same size (2000 tpd) facilities. Among the various differences in assumptions and

other contributing factors, some of the outstanding parameters that have contributed to the high cost

of the 2013 estimate include:

Updated, increased equipment and processing costs based on more recent estimates (e.g.

information from the relatively recent Ensyn/UOP joint venture)

Higher feedstock costs ($88 per metric tonne vs $51 in 2009)

Extra processing steps such as the 3-stage pyrolysis oil hydrotreatment/stabilization process.

Higher Hydrogen consumption to hydrotreat raw pyrolysis oil to a stable hydrocarbon oil of <2

wt% Oxygen (5 wt% Hydrogen on dry pyrolysis oil for 2009 vs 5.8 wt% for 2013)

Much higher indirect costs (engineering, risk, contingency etc.) which represent ca. 100% of

total installed capital (TIC) costs (vs.ca. 50% in the 2009 study)

Two other recent studies (T. R. Brown et al., 2012; Wright et al., 2010) have also assessed the potential

cost of producing pyrolysis biofuels for transportation. The 2010 study, which assumed imported

hydrogen, estimated a similar cost of $2.11/gal (0.56/L) to upgrade the pyrolysis oil, similar to the 2009

PNNL study (after normalizing for the higher feedstock cost assumption). The assumptions in the 2010

study were updated, based on the world’s first commercial pyrolysis facility completed by KiOR, to

reflect some of the characteristics of this new facility. Although the details of the KiOR process are still

not fully disclosed, the authors based their analysis on the preliminary information outlined in KiOR’s

initial public offering document which contained information on the expected capital and operating


July 2014 104

costs of the facility. One figure that had changed significantly was the capital cost which had more than

doubled (from $200 m to $429 m) once the 2012 KiOR details were entered into the 2010 model. A

major contributor to this difference was the power generation equipment cost which, as shown in

Table 3-5, was about 5 times that of the 2010 value (34 vs 141). This implies that KiOR places

considerable emphasis on maximizing power generation for the process itself as well as for export to the

grid. The reasons behind this strategy are likely related to US RFS2 compliance and the GHG emission

savings achieved with increasing amounts of renewable power generated on site at the biofuel facility.

The sale of power to the grid should contribute about $12 m annually to the financial returns of the

facility. This revenue explains why, compared to the 2010 analysis, the MFSP in the 2012 study is only

about 20% higher (2.57 vs 2.11 USD/gal) when the capital cost is 115% higher. Other contributors to the

2012 study’s high capital costs are primarily due to equipment cost inflation, a costlier hydroprocessing

facility (with more emphasis on H2 recycling) and the purchase of refiners to make actual diesel and

gasoline rather than a mixture of liquid hydrocarbons.

In terms of sensitivity analysis, all four studies agree that the MFSP (Minimum Fuel Selling Price) is very

sensitive to both pyrolysis yields and hydroprocessing yields. These sensitivities are indirectly related to

the cost of feedstock (good yield may be irrelevant if cost of feedstock is too high) and the

hydrotreatment catalyst and hydrogen cost (which largely determine the yield of the hydroprocessing

facility). For example, in the 2012 ISU study, dropping the bio-oil yield (Table 3-5) from 63% to 39%

increased the MFSP from $2.57/gal ($0.68/L) to $3.32/gal ($0.88/L).

Table 3-5: Recent techno-economic studies on pyrolysis-derived biofuels

2009 PNNL (own

hydrogen)

2009 PNNL

2010 ISU

study

2012 ISU study (KiOR

based)

2013 PNNL (based on expected

technology by 2017)

Cost year basis 2007 2007 2007 2011 2011

Facility fuel output, MGPY (MLPY)

43 (163)

76 (288)

58.2 (220)

57.4 (217)

60.5 (230)

Biomass to Bio-oil yield (mass) 65% 65% 62% 63% 64%

Biomass to HDO yield (mass) 44% 44% 41% 42% 44%

Biomass to drop-in biofuel yield (mass)

29% 29% 25% 26% 28%

Total Project Investment (million $) 303 188 200 429 700

TPI Hydroprocessing (million $) 110* n/av 27 90 230*


July 2014 105

TPI Fast pyrolysis (million $) 92 n/av ca.30 ca. 20 279

TPI Hydrogen generation (million $)

86 n/ap n/ap n/ap 119

TPI Power generation (million $) n/ap n/ap 34 141 n/ap

Hydrogen cost ($/kg) n/ap 2.2 1.3 1.3 n/ap

Electricity price ($/kWh) 0.064 0.064 0.054 0.054 0.069

Feedstock cost ($/MT dry basis) 51 51 83 83 88

Transportation fuel MFSP, $/gal ($/L)

2.04 (0.54)

1.74 (0.46)

2.11 (0.56)

2.57 (0.68)

3.39 (0.90)

(Jones et al. 2013.,T. R. Brown et al., 2012; Jones et al., 2009; Wright et al., 2010)

*contains product separation and finishing equipment

As for capital costs, hydroprocessing and power generation equipment are the highest expenses and

they must be weighed against the potential benefits of integrating the facility with a refinery and selling

electricity to the grid. The 2010 ISU study also contrasted the capital cost of a first of kind with an nth

plant facility. A pioneer plant would cost $911 million ($585 million for purchased Hydrogen) compared

to a mature plant of $287 m ($200 m for purchased hydrogen scenario). This highlighted the high

equipment risk involved in pioneer pyrolysis facilities, especially those that will produce hydrogen on

site.

The three earlier techno-economic studies (Jones et al. 2009, Wright et al. 2010, Brown et al. 2012)

generally agree on various points including that the cost of producing pyrolysis diesel would be in the

vicinity of $2.5/gal ($0.66/L) which is equivalent to about $19/GJ. This value is similar to the one

calculated by the BIOCOUP group described earlier ($18/GJ) and indicates that upgrading to drop-in

biofuels adds about $10/GJ to the $10/GJ base cost of making bio-oil. However, it must be noted that

the recent and most updated study by Jones et al. (2013) shows a significantly higher cost of producing

drop-in biofuels from a fast pyrolysis platform with the capital cost as TPI reaching $700 million and the

production costs as MFSP climbing to $3.39/gal ($0.90/L) or ca. $26/GJ. Again, these are rough

comparisons and are only meant to give a “high level” sense of the currently estimated costs of pyrolysis

derived drop-in biofuels. However, they can prove to be useful in helping identify the challenges and

opportunities for cost reduction and eventual commercialization of the pyrolysis drop-in platform.

3.6 Pyrolysis biofuel commercialization

Although there are several groups in the process of commercialising pyrolysis derived biofuels, by way of

examples, the three pyrolysis biofuels commercialization efforts of Envergent, Dynamotive and KiOR are


July 2014 106

described below. Each of these companies is pursuing a different technology with Envergent using a

circulating fluidized bed (CFB), Dynamotive a Bubbling Fluidized Bed (BFB) and KiOR using catalytic

pyrolysis (CP):

Dynamotive has been developing its Bubbling Fluidized

Bed pyrolysis technology for several years. The company

has constructed 4 installations to date with the West

Lorne (2006, 130 tpd, no longer active) and Guelph facilities (2008, 200 tpd no longer active) being the

largest (Bradley, 2009). Dynamotive has also been developing its so called BINGO approach, a

proprietary 2-stage bio-oil hydrotreating technology which produces: 1) a liquid intermediate that is

blendable with petroleum feed after the first stage, and 2) a finished transport fuel blend stock after the

second stage. The yields claimed at bench scale are 25 wt% biomass to transport fuel stock (or 300

L/odt) with less than 0.01% oxygen and an even product distribution (20% heavy VGO, 30% Gasoil, 30%

Jet and 20% lights) (Dynamotive, 2009). The company has not disclosed the type of catalysts it is using.

In 2011, Dynamotive signed an agreement with French innovation center IFPEN and Canadian

technology provider Axens to upscale and commercialize the BINGO process. Dynamotive has

emphasised the value of its agrichar and biogas co-products (Dynamotive, 2009).

Envergent is a joint venture between UOP Inc and Ensyn.

Ensyn is contributing its 25 years of experience in

producing bio-oils (circulating fluidized bed CFB reactor

development) while UOP is contributing its expertise in

developing hydrotreatment catalysts and reactors for the

oil industry. Envergent is building a demonstration facility at the Tesoro refinery in Kapolei, Hawaii with

the help of DOE’s US$25 m. The facility will process Ensyn RTP bio-oils and is expected to start operating

in 2014. The plant aims to produce 4 barrels per day of gasoline diesel and jet fuel. This project is part of

a larger vision to install 4 RTP bio-oil facilities and one central upgrading facility which will use UOP

hydrotreating technology. Nine more new plants are being planned in Malaysia in order to use palm oil

residues which according to recent estimates amount to about 150 million dry tonnes annually across

the country (Aziz & Dato’ Leon, 2012). In October 2012, Ensyn forged a strategic alliance with the


July 2014 107

Brazilian Pulp & Paper company Fibria Celulose S.A. (NYSE: FBR). The alliance includes a joint venture

between the two companies as well as US$ 20 million equity investment in Ensyn Corporation by Fibria

(Ensyn, 2013b).

KiOR is one of the Khosla Ventures-backed biofuel start-ups. In April

2011, the company filed for a $100 m initial public offering (IPO) with

Credit Suisse, UBS and Goldman Sachs as underwriters (KiOR, 2013;

Lane, 2013). The plant (capital cost of $190 million) could be

considered the world’s first, commercially operational, biomass derived, drop-in biofuel facility. The

Columbus based facility has the capacity to process 500 ton per day (tpd) of wood (primarily yellow pine

from the local wood basket) into 49.2 MLPY (13 MGPY) of liquid biofuels. A second, larger (3x) facility is

being built in Natchez, which is projected to cost about $350m by the time it is complete. The company

goal is to build a total of 5 commercial facilities at a total of $1 billion in investments. KiOR has indicated

that they will commit $500 million while the state of Mississippi will commit $75m in interest-free loans

(KiOR, 2013; Lane, 2013). KiOR’s technology (as has been discussed previously) consists of catalytic

pyrolysis in a Circulating Fluid Bed configuration followed by hydrotreatment to upgrade to transport

fuel blendstock. The techno-economic analysis and the fact that the KiOR facility has been built in close

proximity to a natural gas pipeline, imply that the strategy might be to use imported hydrogen and all

extant biomass for power generation. Prioritizing power generation will likely provide the facility with

high returns from electricity sales and high GHG emission saving per volume of biofuel produced (T. R.

Brown et al., 2012; KiOR, 2013; Lane, 2013). The company claims that its drop-in biofuel results in a

reduction of 80% GHG emissions when compared to petroleum fuels (KiOR, 2013). Recently (March 17,

2014), KiOR announced a net loss of USD $347.5 million for 2013 as well as bringing the Columbus

facility to an idle state. The company also expressed concerns about whether it would be able to attract

further investor funding and avoid bankruptcy. As KiOR was considered a leader in the

commercialization of drop-in biofuels, this development is a concern for the drop-in biofuel sector.

Major technical problems were listed as the main reasons for the company not being able to meet

performance milestones (Lane, 2014).


July 2014 108

There are a number of other pyrolysis diesel commercialization efforts underway in the USA (such as

Anellotech, GTI’s new IH2 technology) and Europe (KIT and BTG). However, most of these facilities are at

the less-than 25 tpd scale (Meier et al., 2013).

In summary, the production of bio-oil intermediates via pyrolysis is relatively inexpensive and it can be

decoupled from the subsequent upgrading steps thus facilitating feedstock delivery logistics (e.g. bio-oil

is more energy dense and should be cheaper to transport than the original biomass feedstock).

However, the resulting bio-oil is physicochemically disparate from petroleum liquids and thus the

greatest costs and technological challenges of this platform are in upgrading the bio-oil to petroleum-

equivalent transport fuels. Upgrading costs are estimated to account for about two-thirds of the capital

expenses and about half the operating expenses. Hydroprocessing has many advantages as an upgrading

technology, not least among which is that it can efficiently elevate the H/C ratio of the bio-oil prior to

final conversion to transport fuels. Oil refinery FCC and hydrotreating units can be leveraged by bio-oil

upgrading processes, although the insertion of insufficiently deoxygenated bio-oils poses numerous

technical challenges (coking, catalyst deactivation, corrosion) and will cause a reduction in the H/C ratio

of the refinery’s products. However, the significant capital cost savings as well as some synergistic

benefits (coking and viscosity reduction) warrant the continued interest in co-processing partially

upgraded bio-oils with petroleum feeds inside oil refineries. Recent work in the pyrolysis area has

suggested that the greatest benefits in upgrading bio-oils are not necessarily in maximizing oxygen

removal but rather in first removing the most unstable oxygenated compounds by way of a mild

hydrotreatment step. The technical area that could have the biggest impact on facilitating improved bio-

oil upgrading, as implied by the assumptions in the various techno-economic models, is increasing the

selectivity and lifespan of the hydrotreating catalysts that are used.

3.7 Gasification

Gasification, as the name suggests, is the conversion of solid biomass to a gas, with small amounts of

liquids and char also co-produced. The process is typically conducted under conditions of high

temperature and pressure using air, oxygen or steam as a gasifying agent to convert biomass to a low to

medium energy gas known as producer gas or “syngas”. Unlike raw biomass, syngas is relatively

homogeneous and it is comprised of mostly hydrogen and carbon monoxide as well as small amounts of

CO2, H2O and CH4. When air is used as a gasifying agent, 50 wt% of the gas is nitrogen and thus air


July 2014 109

gasification is only appropriate for energy applications while oxygen and steam gasification are

appropriate for the production of synthesis gas. The exact composition of the gas will vary depending on

the composition of the feedstock but mostly on the gasification process conditions that are used.

Although syngas can be used for the same applications as natural gas, it is a more oxygenated gaseous

fuel with less than half the energy density of natural gas (natural gas contains about 36 MJ/Nm3 whereas

biomass syngas contains only about 4-18 MJ/Nm3 (4 MJ/Nm3 for air blown and 18 MJ/Nm3 for steam

blown gasification (Bain, 1992)). The quality of a syngas for synthesis applications is often measured by

its H2/CO ratio. A higher ratio typically indicates a greater energy density and therefore better potential

for upgrading to drop-in biofuels. In contrast to combustion, which results in the conversion of biomass

to thermal energy and fully oxidised gases (CO2 and H2O), gasification takes place under conditions

where oxygen is limited. Therefore, some of the biomass energy is retained in the partially oxidized

gaseous product.

Gasification is not a new process as it has been used since the early years of industrialization. Between

the 1940s and 1970s, in most European countries and the US, “town gas” produced by the gasification

of coal was used to fuel street lamps and cooking stoves. During the Second World War, in Germany coal

syngas was catalytically condensed to liquid transportation fuels (widely known as the Fischer-Tropsch

process). As mentioned previously, the South African company Sasol uses a variation of this technology

to produce about 160,000 barrels per day of coal-derived diesel fuels and chemicals.

3.7.1 Gasification process

The gasification of biomass can generally be viewed as proceeding in four main sequential steps:

1. Drying: moisture is removed from the biomass particles

2. Devolatilization: as the dry particle is heated, it devolatilizes and the resulting volatiles exit the

particle and come into contact with oxygen and other gases in the reactor.

a. This devolatilization step is also known as the “pyrolysis” step (Bain and Broer, 2011)

3. Combustion: After contacting oxygen (O2 or steam), the volatilized carbon is converted to

carbon oxide gases (CO and CO2). An exothermic reaction occurs which, if sufficient oxygen is

present, provides enough heat for the last reduction step.

4. Reduction: This step converts the carbon and carbon oxides to the main components of syngas,

i.e. H2 and CO. Four main reactions take place during this step:

o Water Gas C + H2O → CO + H2


July 2014 110

o Boudouard C + CO2 → 2CO

o Water-Gas-Shift CO + H2O → CO2 + H2

o Methanation CO + 3H2 → CH4 + H2O

The chemical composition of the produced syngas depends on the relative prevalence of the reactions

taking place at the reduction step of gasification. For example when adding steam to the reactor, the

water-gas-shift (WGS) reaction is favoured and more hydrogen is generated which boosts the H2/CO

ratio of the resulting syngas. Alternatively, when hydrogen is fed to the reactor, the methanation

reaction is favoured and the generated syngas is rich in CH4 (useful for the production of synthetic

natural gas) (Bain & Broer, 2011).

3.7.2 Gasifier types

As was described previously for pyrolysis, gasification relies on reactor configurations that maximize

heat transfer through the biomass particles. Although the reactors used for gasification are similar to

the ones used for pyrolysis, they typically operate at higher temperatures (800 – 1000 °C) and pressures

from 1 to 50 bar). The three types of reactors that are best suited for biomass gasification are: fixed bed,

fluidized bed and entrained flow (Swanson et al., 2010). While gasification reactors can be both top- and

bottom-fed, the following discussion will assume top-fed reactors in order to describe and compare

them on a consistent basis. The fixed bed reactor is the simplest and most established design and it

tends to be found in older, smaller scale systems. In these reactors the biomass is fed from the top of

the reactor to form a bed close to the bottom of the reactor. To enhance the biomass particles

gasification, air (or pure oxygen) is blown through the biomass bed. This creates distinct temperature

zones within the reactor with the lower temperature zones closer to the top of the reactor where the

biomass is dried and devolatilized soon after it enters the reactor. The bottom of the reactor, where the

air is blown, is the hottest section of the reactor and this is where the biomass gets partially oxidized

(combustion) before the resulting gases undergo reduction. The gasifying biomass bed sits on top of a

moving grate which removes the residual ash and char solids out the bottom of the reactor vessel. The

final gas mixture is usually recovered out of the side of the reactor. There are two types of fixed bed

reactors and, depending on whether the gases are blowing countercurrent or concurrent to the biomass

feed, they are called updraft or downdraft fixed bed reactors (see Figure 3-8). Fixed bed reactors are

typically quite simple to build and operate. However, their use is limited by poor heat and mass


July 2014 111

transfers, mainly due to the formation of preferential channel flows within the fixed biomass bed (R.

Swanson et al., 2010).

Figure 3-8: Schematic of updraft (a) and downdraft (b) fixed bed gasifiers

Source: (Swanson et al. 2010)

By increasing the volumetric gas flow of the air blown through the fixed bed, turbulence agitates the

particles of the bed (thus ‘fluidizing’ the bed) and the reactor becomes a fluidized bed gasifier. In

fluidized bed gasifiers the bed particles are composed of gasifying biomass as well as of very hot and

small inert particles which greatly improve the heat transfer throughout the gasifying biomass particles.

There are two types of fluidized bed reactors, the Bubbling Fluidized Bed (BFB) and the Circulating

Fluidized Bed (CFB) reactors. These types of reactors have been described earlier in the pyrolysis section.

The difference is that, for effective gasification, the operating temperatures are much higher than are

used for pyrolysis since the objective is to maximize production of permanent gases as opposed to

condensable vapors. As is the case with pyrolysis, there are cyclones at the exit of the reactors in order

to capture fine char particles that are entrained in the produced gases. Compared to fixed beds,


July 2014 112

fluidized beds generally have higher carbon conversion efficiencies and they tend to be more easily

scaled-up (Bain & Broer, 2011; R. Swanson et al., 2010).

An entrained flow reactor is another type of gasifier where biomass is fed in the form of very fine

particles (or atomized pyrolysis liquids) which are entrained into a high-velocity stream of air or oxygen

(Bain & Broer, 2011). These streams are fed from the top of the hot reactor where the temperatures are

high enough (about 1300 °C) that the entrained biomass particles are gasified before they reach the

bottom of the reactor. This temperature is also high enough to melt most of the ash components of the

biomass which form a liquid slag that flows down the inside walls of the reactor. This effectively protects

the metallurgy from corrosive gases. The liquid slag flows to the bottom of the reactor where it is

collected. Limestone is often added as a fluxing material to help form the liquid slug although, for alkali

ash-containing herbaceous biomass, this may not be necessary (Bain & Broer, 2011; R. Swanson et al.,

2010). Entrained flow (EF) gasifiers have been developed for use with coal by companies such as Shell,

Texaco, Conoco Philips (Bain & Broer, 2011) and, so far, they have had limited application to biomass.

One of the reasons is that feed preparation costs are much higher for biomass than they are for coal and

their use also involves extensive drying and size reduction (<1mm) (R. Swanson et al., 2010). Drying is

important as water can act as a heat sink and compromise control over temperature levels in the

reactor. Size reduction is also important, particularly when no bed is formed and the residence time is as

short as the time it takes for a particle to fall the length of the reaction zone. An alternative strategy to

preprocessing biomass is the approach taken by companies such as Germany’s KIT (“bioliq” technology).

They propose feeding EF gasifiers with pyrolysis slurries as opposed to comminuted biomass. The

company claims that this approach results in an improved feed for entrained flow gasifiers and also

offers the option to decouple the conversion of biomass to bioslurries from the central gasification plant

in terms of space, time and synchronization. This concept is briefly described in the pyrolysis section and

it is schematically represented in Figure 3-9.


July 2014 113

Figure 3-9: Schematic of a biomass pretreatment via fast pyrolysis followed by an entrained-flow gasifier Source: (A. van der Drift et al., 2004)

3.7.3 Syngas cleanup

The “clean up” of syngas is an evolving area of gasification (Dayton et al., 2011). Raw biomass syngas

(also termed “biosyncrude”) is recovered along with numerous impurities such as small char particles,

tar vapors as well as volatile nitrogen and sulfur compounds. Char is entrained in the syngas and it is

comprised of non-volatilized biomass as well as ash. The tar component is formed during the

polymerization of biomass vapors and it can stick to reactor walls and catalysts causing clogging and

deactivation. (As described previously; coke formation during pyrolysis oil production). Sulfur and

nitrogen gases are derived directly from the biomass feed and those components are deleterious to

downstream processes as they cause NOx SOx emissions upon combustion and they can also “poison”

the Fischer-Tropsch catalysts (Dayton et al., 2011).

When the syngas is simply burned, these impurities are of lesser concern. However, when the syngas is

used for more “sophisticated” applications such as internal combustion (IC) engines, gas turbines and

Fischer-Tropsch (FT) synthesis, exhaustive cleanup is required. For internal combustion (IC) engines to

operate effectively, particulates, tars and acids must all be at concentrations below 50 mg/Nm3. For FT

Biomass Drying Flash

pyrolysis

Oil and char

slurry

production

Slurry pump Slagging entrained flow

gasifier 4MPa, 1300-

1500oC

Cooling and

power

generation

Syngas clean up

and conditioning Syngas


July 2014 114

synthesis the particulate matter must be less than 0.03 mg/Nm3. Unfortunately, for most raw syngases

the contaminant concentrations typically exceed these limits by some orders of magnitude (see Table 3-

6). This is especially true for fluidized bed gasifiers which, when compared to entrained flow gasifiers,

operate at lower temperatures and circulate the gas through beds that are rich in fine char solids. As

summarised in Table 3-6, a Circulated Fluid Bed (CFB)-derived syngas typically contains about 10,000

mg/Nm3 of either char or tar particles, Bubbling Fluidised Bed (BFB) up to 43,000 mg/Nm3 and

downdraft fixed beds as much as 30,000 mg/Nm3.

Table 3-6: Char and Tar content of biomass syngas from different reactors

Reactor type Tar content (mg/Nm3) Char content (mg/Nm3)

Updraft 50,000 nd

Downdraft 1,000 9,300 – 30,000

BFB 10,000 1,040 – 43,610

CFB 10,000 1,700-13,100

Specification for IC engines <50 <50

Specification for FT synthesis <0.02 < 0.02

Source: Data from (Bain & Broer, 2011; Carpenter et al., 2010; A. Van Der Drift et al., 2001; Meehan,

2009; Milne et al., 1998; Wander et al., 2004)

There are various ways to reduce the accumulation of these contaminants during gasification but they

all require trade-offs. For example, at high gasification temperatures (e.g. >1200 °C) the formation of

char and tar is reduced and the production of permanent gases is favoured. However, high temperature

gasifiers are costly to build and operate. Similarly the introduction of steam has been shown to reduce

the buildup of tar by reforming it to H2 and CO. However, steam can act as a heat sink and it can

significantly compromise the heat balance and overall efficiency of the gasification plant. Higher air to

fuel ratios cause more oxidative conditions and they are beneficial in reducing the char and tar content

of syngas but these conditions favour more full oxidation of gases to CO2, thus recovering less

combustible gases such as CO. Pure oxygen can be used instead of air in order to improve the specificity

of the gasification reaction while also reducing the nitrogen contaminants coming from atmospheric air.

However, the isolation and purification of oxygen weighs significantly on the economics of the facility

(Bain & Broer, 2011; Muellerlanger et al., 2007; R. Swanson et al., 2010). Clearly, a compromise has to


July 2014 115

be found between these trade-offs in order to strike the optimal quality syngas for each given

application.

The ash content, another significant contamination in syngas, is largely dependent on the amount of

inorganics in the initial biomass. As noted earlier, herbaceous and agricultural biomass tends to be

higher in inorganics than wood derived biomass. Alkaline components of ash such as calcium and

potassium are particularly undesirable since they are known to lower the overall melting point of

biomass, thus promoting slagging at lower temperatures (Hayes, 2013; R. Swanson et al., 2010).

Even if the process is optimised to minimize syngas impurities by adjusting gasification conditions, some

cleanup is always needed. Syngas cleanup is one of the most expensive steps of a gasification biofuel

platform. This typically involves various sequential steps that follow the gasification reactor including gas

cyclone removal of most of the particulate matter above 10 μm and further removal of the smaller

particles by more costly methods such as wet scrubbers or electrostatic precipitators (ESPs) (Dayton et

al., 2011).

Tars are particularly problematic for Fischer-Tropsch (FT) fuel synthesis as they can irreversibly

deactivate the catalysts downstream. One of the reasons for the recent Range Fuels biofuel facility

closure was that the gasification operations resulted in too much of a tar build up (Hayes, 2013; Lane,

2012)). Tars are typically removed by one of the two main strategies of wet scrubbing at low

temperatures or cracking at high temperatures. Wet scrubbing involves significant cooling of the gas as

well as the generation of wastewater streams that are contaminated with potentially toxic tar

components which must be treated prior to disposal. As hot gas cleaning occurs at temperatures higher

than the gasification temperatures (e.g. 1200 °C) the added energy requirements, including the need to

extensively cool the gas down prior to FT processing, can compromise a facility’s economic viability (G. S.

Speight & Ancheyta, 2007). Another cleanup option that has attracted research attention lately is the

catalytic conversion of the tars in the gasifier outlet via steam reforming (Bain et al., 2014). The primary

objective of tar control strategies is to either eliminate tars or to at least bring their dew point below the

operating temperatures of the ensuing processes such that tar deposition is minimized. For more detail

on the various syngas cleanup and conditioning technologies proposed, the reader is referred to a

recent comprehensive review by Dayton et al. (2011).


July 2014 116

Gasification yields tend to be lower than desired as much of the original carbon is oxidized to generate

heat. Only about half of the carbon is converted to usable CO while the other half is oxidized to CO2

(Hayes, 2013). However, it has been recently demonstrated by researchers at the Universities of

Massachusetts and Minnesota that the addition of methane to the gasification reaction can substantially

enhance gasification yields. This is thought to be a result of methane having a higher H/C ratio of 4, thus,

the hydrogen from the methane reacts with the CO2 to form H2O and CO (Hayes, 2013).

3.7.4 Syngas uses

The uses of syngas are similar to those of bio-oil in that they are mostly used for power and heat. The

technological challenges in the utilization of syngas can be ranked as; burners < gas internal combustion

(IC) engines < synthesis to biofuels and chemicals. Burners are robust but do not provide as much energy

efficiency as IC engines. Synthesis has, by far, the lowest energy efficiency but it yields high value-added

biofuels and chemicals. A popular power generation configuration for gasification facilities is the

integrated gasification combined cycle (IGCC). This configuration is designed for maximizing power

generation and it utilizes a high-efficiency gas turbine to burn the clean gas while the exhaust heat in the

flue gases is recovered through a steam turbine that contributes additional power generation (Craig &

Mann, 1996).

For those IGCC engines which combine heat and power recovery, power efficiencies of 35-40% have

been reported and it has been suggested that efficiencies of 50% can be achieved (Craig & Mann, 1996).

Although there are numerous commercial facilities generating heat and/or power from biomass

gasification most of these facilities are smaller scale, in the range of 2 to 50 MW. According to the US

National Energy Technology Laboratory (NETL) world gasification database, a total of 9 biomass

gasification plants were operational in 2010, generating a total of 373 MWth (NETL, 2013). This does not

include the contributions of more recent biomass gasification plants or co-gasification facilities.

3.8 Syngas catalytic condensation for synthesis

As mentioned above, a higher-value application than just burning the syngas is to “condense” the gas to

hydrocarbon liquids that can then be used for the synthesis of biofuels and chemicals (Dayton et al.,

2011). This condensation has to be selective and produce the targeted hydrocarbons, thus, specialized

catalysts must be employed. The types of fuels and chemicals that syngas can be condensed to are

http://www.netl.doe.gov/technologies/coalpower/gasification/worlddatabase/


July 2014 117

summarised in Figure 3-10. These catalytic condensation reactions convert the syngas molecules (H2 and

CO) into larger molecules to produce energy dense liquids described in figure 4-3. The choice of

targeted chemicals and the efficiency of each of the conversions depend on the choice of catalyst,

condensation conditions used and the H to CO ratio. More important than all of the above aspects is the

level of contaminants in syngas which must be very low and preferably less than about 10 mg/Nm3.

Sulfur in particular is “poisonous” to FT catalysts and must be at concentrations that do not exceed 2

mg/Nm3 (Dayton et al., 2011). The syngas catalytic condensation pathways that are most relevant to

drop-in transport biofuels are the methanol to gasoline (MTG) and Fischer-Tropsch (FT) processes.

Figure 3-10: Main Syngas Conversion Pathways Source: (Huber et al., 2006)

3.8.1 Methanol to Gasoline process

The first step in this process involves the exothermic conversion of the syngas to methanol, which is a

commercial technology. As long as the syngas is pure and it contains the required ratio of hydrogen to

carbon monoxide, the conversion takes place in the presence of relatively inexpensive Cu/ZnO/Al2O3


July 2014 118

catalysts at temperatures that range between 220-275 °C and pressures of 50-100 bar. Under these

conditions the catalyst can last as long as 2-5 years (Swanson et al., 2010). Although the per pass

conversion efficiency is typically about 25%, Wender (1996) has shown that up to 99% conversion can be

achieved with gas recycle. The core reaction of this conversion is the reaction of one molecule of CO

with two molecules of H2 to form one molecule of methanol.

CO + 2H2 → CH3OH

Methanol is a liquid fuel but not a drop-in transportation fuel. However, it can be turned into a drop-in

gasoline equivalent using the methanol-to-gasoline process (MTG). This process was first developed and

patented by Mobil Oil Corporation (now Exxon Mobil) in the 1970s. The process entails two steps. In the

first step, methanol is dehydrated over an alumina catalyst at 300 °C and 27 bar, to form a mixture of

dimethyl-ether (DME), methanol and water. In the second stage, this mixture reacts over a zeolite

catalyst (359°C and 20 bar) and is further dehydrated to gasoline range hydrocarbons (44 wt%) and

water (56 wt%) (Phillips et al., 2011; R. Swanson et al., 2010).

The efficiency of the process can be improved by stopping the reaction after the first stage since DME

can be used as a truck fuel (high cetane number). Dimethyl-ether is a gas at room temperature so it

would have to be compressed before it is fed to an automobile engine (Phillips et al., 2011; R. Swanson

et al., 2010).

In the 1980’s, Mobil used the MTG process to run a commercial plant in New Zealand of about 14,500

barrels per day of gasoline that was marketed as “M-gasoline”. The facility involved multiple reactors in

order to dissipate the heat that was generated from the highly exothermic reactions. Multiple reactors

also allowed for more regular cleanup of the catalyst beds. The facility was wound down in 1997

because of variable crude oil and natural gas prices and because of ownership issues of the facility.

3.8.2 Fischer-Tropsch process

In 1923 Franz Fischer and Hans Tropsch identified catalysts that can condense carbonaceous gases to

alcohols. The process was patented and named “synthol”. The main motivation for this research was

the scarcity of petroleum fuel in post-world war one Germany. The process was further developed to

produce non-oxygenated liquid hydrocarbons as feedstocks for fuels and chemicals using improved

catalyst formulations. This process became known as Fischer-Tropsch Synthesis (FTS). The process has

undergone continued investigation and optimisation and it has since been adapted for a diversity of


July 2014 119

feedstocks including biomass. More than 4000 publications have been dedicated to the FTS process and

there is a specialized website www.fischer-tropsch.org which focuses on the history, research and

industry developments around this pivotal industrial technology. Fischer-Tropsch Synthesis can use

syngas derived from any source including biomass, coal or natural gas and it can produce precursors for

a wide range of drop-in chemicals and fuels (see Figure 3-10). As long as the syngas is treated and

conditioned properly and there is a good H2/CO ratio, functional and chemical equivalence can be

achieved between the syngas derived from these disparate feedstocks. Although with a biomass

feedstock it is more difficult to achieve the same level of syngas purity as with natural gas, a near

chemical equivalence can be reached. Thus it is possible to scale-up biomass based FT processes by

leveraging the know-how and facilities of existing natural gas and coal gasification FTS plants. The FTS

reaction takes place over specialized catalysts and, as is the case with the previously described MTG

process, FT is essentially a dehydration reaction that it is highly exothermic (R. Swanson et al., 2010).

The basic reaction is represented in the equation below:

CO + 2.1 H2 → - (CH2) - + H2O (delta Hr (227 °C) - 165 kJ/mol

The catalyst surface acts as an “anchor” upon which the carbon monoxide and hydrogen adsorb (see

Figure 3-11 below). The chain growth begins once the carbon monoxide has been broken down,

enabling the coupling of carbon and hydrogen and the separation of oxygen (which leaves as a water

molecule - dehydration). Chain growth continues by adding further CO and H2 until the newly formed

hydrocarbon molecule is desorbed from the catalyst surface.

Figure 3-11: Simplified representation of Fischer-Tropsch chain growth on a catalyst surface Source: (Blades et al., 2005)

http://www.fischer-tropsch.org/


July 2014 120

The pressures used during the FT process range from 10-to-40 bar and the nature of the hydrocarbons

produced is influenced by the temperatures and catalysts used. Higher temperatures (300 – 350 °C) and

iron catalysts produce gasoline, while lower temperatures (200 -240 °C) and cobalt catalysts produce

diesel and waxes. The distribution of heavier and lighter hydrocarbons can be estimated by employing

the Anderson-Schulz-Flory probability model in which longer chain hydrocarbons form as the

temperature decreases. The ratio of H2 to CO also influences the product distribution with high ratios

favouring the formation of lighter hydrocarbons. Iron catalysts favour the Water Gas Shift (WGS)

reaction such that the H2/CO ratio is increasing. With cobalt catalysts, WGS is not favoured and it has to

be performed separately in order to achieve the desired high H2/CO ratio (E4tech, 2009). In contrast to

pyrolysis, the composition of the FT liquid product can usually be predicted from the composition of the

syngas and the reaction conditions used (Dayton et al., 2011). By contrast, as discussed earlier in the

pyrolysis section, the reaction kinetics involved in upgrading pyrolysis oils cannot be readily modelled,

primarily because the composition of the bio-oil itself is very complex and hard to determine

analytically. As the liquid hydrocarbons produced from FTS are very low in aromatics and transport fuel

blends usually require some amount of aromatic content, they typically have to be blended with

aromatic petroleum cuts (Bauen et al., 2009).

The choice of catalyst used is important in FTS. As was reported earlier for bio-oil upgrading, the group

VIII transition metal oxides are regarded as good CO hydrogenation catalysts, based on the three main

performance characteristics of lifetime, activity and selectivity. Lifetime largely depends on the quality

of the syngas. Prime quality natural gas-derived syngas catalysts have been reported to last up to 3-5

years. The most active metal catalysts for FTS, ranked in order of activity, are Ru>Fe>Ni>Co. Ruthenium

catalysts are the most active, (as it was for pyrolysis oil upgrading) but they are considerably more

expensive than the iron catalysts (IndexMundi, 2013) and (as noted earlier) about 100 times more

expensive than Nickel and Cobalt. Nickel is a methanation catalyst and does not have the broad

applicability in FTS that other FT catalysts have. Iron has WGS activity but it is an acidic catalyst and it

promotes carbon deposition and coking resulting in lower yields and reduced catalyst lifetimes as seen

in the pyrolysis oil upgrading section. Cobalt is more alkaline and although about 200 times more

expensive it produces higher yields and provides for longer catalyst lifetimes. However, it has low WGS

activity and needs a separate WGS step to boost the H/C ratio of the feed to >2. Finally, as well as being

highly active, a catalyst must also be selective in promoting the reactions that are most desirable. For


July 2014 121

example, Iron/manganese/potassium catalysts have shown selectivity for C2-C4 olefins as high as 90%

(Bain & Broer, 2011).

It is also recognised that temperature control is crucial for effective Fischer-Tropsch synthesis (FTS). The

FTS reaction is so exothermic that the heat generated can irreversibly deactivate the catalyst if it is not

dissipated. Another reason why FTS temperatures must be kept under tight control is that, as the

temperature approaches 400 °C, the methanation reaction is favoured. However, this is only desirable

when the target of the conversion is to make synthetic natural gas (SNG). Therefore FT synthesis reactor

design must be such that the heat from the reactors can be readily dissipated. Various reactor designs

such as tubular and slurry-based have been developed for this purpose (R. Swanson et al., 2010).

Overall, the FTS-from-biomass process involves four major steps (Figure 3-12 ). These steps are; a)

syngas production, b) syngas cleanup, c) FT synthesis and, d) product upgrading. As is apparent from the

schematic in Figure 3-12, each one of these steps is performed at different temperature and pressure

conditions. From an energy balance and economic viability perspective, the high pressure and

temperature fluctuations throughout the process are of some concern. The high pressure consumes

energy and it also results in greater capital expenditures as the gas compressors needed to feed the

pressurized reactor vessels are among the most costly pieces of equipment in a gasification plant (see

subsequent techno-economic section). Similarly, heating and cooling cycles incur undesirable energy

“losses”. If the syngas is cleaned of tars through thermal cracking at >1300 °C, it then has to be cooled

down to about 300 °C and pressurized to >30 bar before it enters the FT synthesis reactors. If, on the

other hand, the tars are removed by way of quenching, they have to be cooled down before they enter

the wet scrubbers and again heated and pressurized before they enter the FT synthesis reactors. All of

these heat-cool and compression operations are costly and they add significantly to the capital cost of

biomass to liquids facilities. There is ongoing research and development to improve gasification

platforms so as to minimize these costly fluctuations in temperature and pressure.


July 2014 122

Figure 3-12: Simplified schematic of the overall biomass gasification to FT diesel and gasoline process

Source: Adapted from (Bain & Broer, 2011)

The final stage of the FTS process (Figure 3-12) is the hydroprocessing of the resulting hydrocarbon FT

liquids. This step can be viewed as equivalent to the previously described processing of highly

deoxygenated HDO bio-oil. These FT liquids are essentially oxygen-free and they can be easily upgraded

(hydroprocessed) inside an existing petroleum refinery. For example, the Karlsruhe Institute of

Technology (KIT) has proposed the co-location of its bioliq® process (a “hub and spoke” pyrolysis-then-

gasification approach) with an oil refinery (Dahmen et al., 2012). The core gasification-FT facility would

be located next to a petroleum refinery where the last hydroprocessing step would be performed

without the need to build separate hydroprocessing reactors.

The Fischer-Tropsch process is currently being operated at an industrial scale by two main fossil fuel

companies in South Africa (Sasol) and in Malaysia and Qatar (Shell). The world’s first commercial-scale

gas-to-liquid (GTL) plant based on FT synthesis was completed in 1993 by Shell in Bintulu, Malaysia and

continues producing about 15,000 barrels per day (bpd). The previously discussed Sasol plant produces

Syngas

production

Gas cleanup

and

conditioning

Clean syngas

H2/CO

Low temp FTS

Cobalt catalyst

High temp FTS

Iron catalyst

Wax (c>20)

Hydroprocessing

Olefins (C3 – C11)

Hydrogenation

diesel

gasoline Biomass Coal

Natural gas

Air Oxygen Steam

Up to 30 bar Up to 150 bar

Up to 350 °C Up to 400 °C

Cool down or overheat

Sulfur Removal Water Gas Shift

Particulate removed Thermal treating of tar Wet scrubbing of tar


July 2014 123

160,000 barrels of FT-diesel per day from coal syngas to provide 41% of South Africa’s transport fuel

requirements. Sasol has also converted one of its coal-to-liquid (CTL) facilities to accept natural gas from

Mozambique. In December 2006, Sasol started up its 34,000 bpd Oryx gas-to-liquid (GTL) facility in Qatar

and today this facility produces more than its design capacity. Sasol is using the experience gained from

building and operating this facility to expand operations in other countries which are endowed with

significant natural gas reserves such as Nigeria, Uzbekistan, the USA and Canada (Sasol, 2013).

The world’s largest GtL facility was built by Shell in collaboration with Qatar Petroleum. It has a capacity

of 140,000 bpd and it has been fully operational since 2012. Construction of this facility began in 2007

with an original timescale of 2 years and a budget of $5b. Five years later the facility was finally

completed but at a final cost of about $19 b (Platts, 2013). There were several difficulties encountered

with the conversion of natural gas to liquid hydrocarbons and some problem areas have yet to be fully

resolved. Thus, it is likely that a biomass FT gasification process will encounter similar or even more

challenging problems. It has been suggested that when a biomass-to-liquids process is compared to a

GtL plant the main challenges will be: processing a more heterogeneous biomass feedstock, producing a

lower quality syngas and the smaller scale and feedstock availability risks (Hileman et al., 2009).

3.9 Syngas upgrading techno-economic considerations

The cost of gasification-derived biofuels can be estimated quite accurately since the processes are based

on established industrial processes in the coal gasification industry. A recent techno-economic study by

ISU/ConocoPhillips/NREL (Swanson et al., 2010) has estimated the cost of gasoline produced from FT

conversion of biomass syngas. The study assumed a corn stover cost of $83/odt, a 2000 tpd facility and a

10% ROI and compared two scenarios of: a) low temperature using a fluidised bed gasifier at 870 °C, and

b) high temperature using an entrained flow gasifier at 1300 °C. Both scenarios were followed by

Fischer-Tropsch synthesis and hydroprocessing of the resulting FT liquids. Some of the results for both

the high temperature (FTHT) and the low temperature (HTLT) scenarios are summarised in Swanson et

al., (2010). Although the FTHT scenario had 20% higher capital costs, its minimum fuel selling price

(MFSP) was about 10% lower. This was achieved because of the much higher conversion efficiency of

the high temperature approach. The highest sensitivities of the FTHT scenario are mainly the capital

costs, feedstock costs and compressor capacity. A 30% increase in capital costs resulted in an 18%

increase in MFSP. An increase in the feedstock cost from 75 to $100/odt, resulted in a MFSP of 10%


July 2014 124

increase. An increase in the compressor install factor from 1.2 to 3 further increased the MFSP by 17%

(Swanson et al., 2010).

Table 3-7: Techno-economic studies on gasification-derived biofuels (plant size: 2000 tpd of dry basis biomass, cost basis year: 2007)

2010

NREL

(FTHT)

2010

NREL

(FTLT)

Facility fuel output (MGPY)

(MLPY)

39.8

(150)

32.3

(122)

Fuel yield (gasoline gallon Equivalents/metric tonne)

(gasoline L Equivalents/metric tonne)

61

(231)

47.2

(179)

Total Project Investment (TPI, million $)

(equipment only, million $)

606

(309)

498

(254)

TPI for Gasifier (million $) 68 28

TPI for Syngas cleanup & conditioning 34 29

TPI fuel synthesis 49 59

TPI Hydroprocessing 33 30

TPI Power generation 46 39

Transportation fuel MFSP (Minimum Fuel Selling Price ($/gal)

($/L)

($/GJ)

4.27

(1.13)

(31.4)

4.83

(1.28)

(35.5)

MFSP Feedstock 28.9% 32.9%

MFSP Capital Depreciation 14.8% 13.9%

Feedstock cost ($/MT) 75 75

Source: (Swanson et al., 2010)

Overall, the estimated MFSP and capital costs for both low and high temperature scenarios are high and

they range between $4.3 and $4.8/gal ($1.1-1.3/L or $31-36/GJ), and between $500 and $600 (TPI)

million respectively. These costs represent mature facilities while pioneer facilities were estimated to

have about twice the TPI capital cost ($1400 million for the high temperature scenario and 830 million

for the low temperature scenario). These costs are double the values estimated for pyrolysis, as


July 2014 125

discussed earlier. However, a direct comparison is difficult to make, both because of the known

disparities in assumptions and costing of these types of analyses and also because of technological

maturity differences. Despite its high apparent cost, the gasification platform can be viewed as being

more technologically mature than pyrolysis (especially when compared to the low maturity of pyrolysis

oil upgrading technologies) and it has also leveraged know-how from industrial coal gasification. For

example, the assumed catalyst lifespan in gasification is about 3 years, and it is more likely to be reached

than the 1 year catalyst lifespan projected for pyrolysis oil upgrading as discussed in Section 3.5.

When the sensitivities of the pyrolysis and gasification analyses are compared, the biggest sensitivities

for pyrolysis relate to yields, catalysts and hydrogen costs while for gasification they are the capital

expenditure and the feedstock costs. Compared to pyrolysis, gasification is a high capital expense

platform due to the higher operating temperatures that are needed, the complexity of the process, the

multiple process steps and the requirement for various heat-cool and compression cycles. As gas

compressors are very energy and capital-intensive the compressors accounted for about 18% of total

equipment costs. The compressor capacity of a gasification facility is directly related to the process

efficiency, thus the lower the per pass yield of a process, the more a compressor has to work to recycle

the gas stream enough times until it has reacted fully.

When comparing gasification to pyrolysis facilities, hydroprocessing equipment (mainly used for

cracking the heavy ends of FT liquids) usually represents a lower proportion of overall capital costs (see

Table 3-7). The insertion of hydrogen into a gasification process contrasts with what was discussed

earlier for pyrolysis. In gasification, unlike pyrolysis, there is usually minimal to no external hydrogen

supply (except for some cases like the gasification platform of Sundrop Biofuels which uses natural gas-

derived hydrogen inputs) and the biomass syngas itself is the source of the energy and hydrogen needed

to reduce and deoxygenate carbohydrates to drop-in biofuel hydrocarbons.

Another recent NREL techno-economic study by Phillips et al. (2011) looked at a gasification biofuel

platform based on the methanol-to-gasoline (MTG) process as opposed to the Fischer-Tropsch process.

This study assumed a 2000 tpd facility with 10% ROI and a feedstock cost of around $56/odt (as

compared to $83/odt in the FT studies). The results of the MTG study show that both the capital ($

1.95/gal or 0.52/L) and operating costs ($199 m) were low compared to those calculated by Swanson et

al. (2010) for the FT process. Although these two studies should not be directly compared, these low

cost estimations for the MTG design case warrants further investigation as to whether the MTG


July 2014 126

approach is indeed a lower cost conversion process than Fischer-Tropsch for biomass-to-gasoline

gasification platforms. The main sensitivities of the MTG design were the same as those for the FT

designs with capital cost showing the greatest sensitivity.

As the capital cost showed the greatest sensitivity for the gasification platform, it is not surprising that

large scale facilities are projected to reduce the high MFSP of gasification biofuels. It has been suggested

that a 500 MWth, facility, or one with 2000 tpd of feed or producing 95 MLPY (25 MGPY) of biofuel, is the

minimum scale at which a biomass FT facility can be viable (Hayes, 2013; Chemrec, 2012; NNFCC, 2009).

This would not be considered to be a large facility when compared to the currently operating fossil FT

facilities of Sasol and Shell. (Table 3-8 provides a direct comparison of the scale of these facilities).

Table 3-8: Scale of selected GtL and BtL facilities

Source: Company websites and (R. M. Swanson et al., 2010)

What has been modelled to represent a full scale, biomass-to-biofuel facility is at least 50 times smaller

than the current commercial Sasol CtL (coal to liquid) main gasification facility and even smaller than

Shell’s Pearl GtL (gas-to-liquids) facility. As noted, the scale of a biomass facility will likely be limited by

its access to feedstock and thus a larger scale facility will be challenged to reach the scale of the fossil

based facilities which utilise denser and more transportable feedstocks.

Although gasification is generally viewed as a process requiring large scale facilities to be more

economically attractive, some technology providers have recently advocated the construction of smaller

plants. For example, the company Velocys has built small scale gas conversion facilities to liquefy the

flair gas from ocean oil rigs. Leveraging on this experience, the company is now trying to commercialize

a biosyngas-FT liquids platform based on a microchannel reactor design which can be applied to small

Technology Facility Scale

MLPY(MGPY)

Biomass (Corn Stover) gasification (2000 tpd) ISU/COP/NREL techno-economic

analysis

133 (35)

Coal gasification Sasol, Secunda, South Africa 6,250 (1,650)

GTL (gas to liquids) Pearl GTL, Shell, Natural gas, Qatar 15,900 (4,200)


July 2014 127

scale biomass gasification platforms (around 300-500 tpd). Velocys is also providing the upgrading

technology for Solena Fuels, a FT-biofuel company (Velocys, 2014).

3.10 Progress in the commercialization of biomass gasification and fuel synthesis

Currently there are various commercial-scale gasification facilities that are either operational, under

construction or in the planning stages. While most of these facilities have been built for heat and power

generation, there are several that hope to also manufacture liquid biofuels. There have been and will

continue to be, valuable lessons learned from the more established, commercial heat and power-

generating gasification facilities. Some of these facilities are reviewed below and are listed according to

gasifier type and technology provider.

3.10.1 Fixed Bed gasifiers

As discussed earlier, fixed bed gasifiers are mostly suitable for small scale applications and are not

ideally suited for drop-in biofuel synthesis. Updraft gasifiers are less favoured for drop-in biofuel

synthesis as they typically produce a tar-rich syngas (see Table 3-6). However, there are various

technology developers that are trying to commercialize updraft gasifiers at scales of 2-40 MWth for heat

and power applications. An example of such a facility is the newly completed (2012) Nexterra Inc. 2

MWe updraft gasifier located on the Vancouver campus of the University of British Columbia

(www.nexterra.com). These types of gasifiers are best suited for local small scale CHP applications.

Although many useful lessons can be learned from these facilities, their commercial applications are not

reviewed in detail.

3.10.2 Bubbling Fluidized Bed facilities

The Carbona company (a subsidiary of the Andritz Group), is currently

trying to commercialize the RENUGAS technology which was originally

developed by the Gas Technology Institute (GTI) in Chicago, US. The

technology involves directly heated, pressurized, steam and oxygen-blown BFB gasification coupled with

syngas-fed IC gas engines for CHP generation (E4tech, 2009). Carbona’s technology has been tested in

the Tampere, Finland pilot facility (72 odt/day) since 1993 (E4tech, 2009). A larger 150 odt/day (5.4

MWe, 11.5 MWth) Carbona gasifier (partly funded by the US DOE) has been in stable operation at the


July 2014 128

Skive (a local district heating provider) facility in Denmark since 2012 (Hansen, 2013). The Skive facility

uses three GE Jenbacher (Austria) gas IC engines and it has also recently tested the production of liquid

fuels using the Haldor Topsoe’s Tigas process (a variation of the methane-to-gasoline process) (Hansen,

2013).

Enerkem is a Canadian company that uses its own gasification

technology to produce methanol and ethanol through catalytic

condensation of biomass (or municipal solid waste, MSW) derived

syngas. The Enerkem platform involves a fluidized bubbling bed gasifier followed by cyclone filtration

and wet scrubbing of the syngas. The clean syngas is then catalytically converted to mixed alcohols. A 5

MLPY (1.3 MGPY) demonstration facility has been operating at Westbury, Quebec since 2009, gasifying

old electricity poles to make syngas (since 2009), methanol (since 2011) and ethanol (since 2012). A

second 38 MLPY (10 MGPY) facility has been constructed in Edmonton, Alberta and will use a MSW

feedstock (2014). A third 38 MLPY (10 MGPY) facility is at the licensing stage at Pontotoc, Mississippi,

USA. A fourth 38 MLPY (10 MGPY) facility at Varrennes, Quebec is at the planning stages and it will be a

joint venture between Enerkem and Canadian corn ethanol company Green Field Ethanol Inc. (Enerkem,

2013)

Carbona’s RENUGAS technology is the gasification technology of choice for the

proposed UPM (Finnish UPM-Kymmene Oy) Stracel biofuel facility in France. The Gas

Technology Institute in Chicago is testing the gasification platform for UPM at its 5

MWth pilot plant. UPM has recently received 170 million Euro from the EU’s NER300

program and they expect to begin construction of the 300 MWth facility in 2014

(Landalv, 2013).

3.10.3 Circulating Fluidized Bed facilities

Foster Wheeler (FW) is among the oldest

gasification technology providers. The main

technology is a relatively simple atmospheric circulated fluid bed (CFB) gasifier which was developed in

the 1980s. The first three biomass gasifiers were built in the early 1980s for the European pulp and


July 2014 129

paper industry in order to replace oil –fired lime kilns. These facilities are still operating and they include

a 15 MWth gasifier in Portugal (Portucel’s Rodao), a 20 MWth in Sweden (Norrsundet) and a 28 MWth in

Finland (Schauman mill’s Pietarsaari). More recently the Foster Wheeler CFB technology has or will be

used in four additional facilities:

In 1993, at a 14 MWth IGCC demonstration facility in Varnamo which belongs to Sweden’s utility

company E.ON Sverige (previously Sydkraft). Although this plant has been mothballed since

2011, in 2013 the company applied for 450 million Euro form the EU’s NER300 program to help

build a 200 MWth facility named “Bio2G”.

In 1998, at a 48 MWth facility in Kymijärvi which belongs to Finland’s utility company Lahden

Lämpövoima.

In 2001, at a 32 MWth facility in Varkaus which belongs to Finland’s coreboard company

Corenso. The facility currently produces heat for the lime kiln of the pulp mill and it also

produces small amounts of Fischer-Tropsch waxes.

In 2007, Foster Wheeler was the gasification technology provider for the commercial FT diesel

facility from the joint biofuel venture between Neste Oil and Stora Enso (NSE biofuels Oy).

However, plans to build this facility were shelved in August 2012, after the company failed to

obtain the requested 500m Euro of NER300 funding.

(EBTP, 2013; Foster Wheeler, 2013)

Thyssen Krupp Uhde is a global engineering company that has

long engaged in developing gasification technologies for various

feedstocks including biomass. Among Uhde’s gasification

platforms the most relevant to biomass are the HTWTM (High

temperature Winkler) and PRENFLOTM (entrained flow) reactors. The HTW process is a CFB design which

was developed in the early 1970s by Rheinbraun (now RWE) and it is the pressurized version of an even

older CFB technology which was developed by the German, Fritz Winkler (hence the name) in the 1920s.

The technology has been incorporated into various pilot plants at operating pressures that ranged 1.5-

27 bar, temperatures between 700-950 °C and carbon conversion efficiencies (gasification step only)

that reached 95%. Uhde has recently purchased the HTW technology from RWE and applied it to various

projects. These projects are at the design and development stages in various locations around the world

including Japan, Sweden, India and Australia (E4tech, 2009; Radtke, 2011).


July 2014 130

Uhde’s HTW technology was selected by the

Colorado-based company Sundrop fuels. Sundrop’s

overall technology platform is a methanol to gasoline

(MTG) process which will use natural gas as well as solar heating inputs. Sundrop is a good example of a

process that utilizes natural gas to produce hydrogen-replete biomass syngas. The hydrogen-rich CH4 in

the natural gas is essentially utilized to replenish the hydrogen deficient biomass feedstock and to

increase yields. Solar power is used as a way to increase the renewable energy credit of the facility. A

3500 barrel/day facility in Alexandria Louisiana USA is planned to begin operations in 2014 (Sundrop

Fuels, 2013).

Uhde’s HTW has also been the technology of choice

for the Swedish gasification company

Varmlandsmetanol AB. This well-funded project (recently approved for an additional 11 million Swedish

kroners) hopes to produce 15 MWth and 100 kton/y methanol (total of 111 MWth equivalent) with

operations scheduled to begin in Q4 2015 (Värmlandsmetanol, 2012)..

Another Uhde gasification technology relevant to biomass is the

entrained flow gasification platform called PRENFLOTM. This

platform, originally developed for coal gasification, is well suited for

a torrefied biomass feed and it is at the heart of a fairly recently

announced (2010), 113 million Euro BioTfuel project. The BioTfuel consortium aims at producing FT

drop-in biofuels derived from the gasification of torrefied wood biomass using two planned facilities in

France (Dunkerque and Compiegne locations) by the year 2016 (Radtke, 2011).

3.10.4 Dual Fluid Bed facilities

The Dutch ECN research institute is commercializing a dual-bed CFB steam

gasifier and air blown BFB char combustor (similar concept to Repotec

although reported to be more compact and indirectly heated) technology

which is marketed under the acronym MILENA (ECN, 2008). ECN has also developed an oil-based tar

removal technology which is marketed by the company Royal Dahlman with the acronym OLGA. The


July 2014 131

MILENA and OLGA technologies will be assessed at a 12 MW facility in Alkmaar (The Netherlands) which

is currently under construction and expected to begin operations in 2014 (B. van der Drift, 2013).

Repotec has collaborated with the pulp and paper technology

provider Metso to provide the gasification technology for the

new GoBiGas facility in Goteborg Sweden. Phase 1 of this

ground-breaking facility will produce 32 MWth, in addition to 20 MW methane. The high purity methane

will be produced using Haldor Topsoe technology and it will be fed into a specialized grid catering to a

local fleet of 40,000 gas-powered automobiles. The plant is almost complete and is expected to start

operations by Q4 2013. The facility has also received EU NER300 funding support (Gunnarsson, 2013).

Figure 3-13: The Goteborg biomass gasification to methane facility Source: (Gunnarsson, 2013)

Repotec / TUV (Vienna University of Technology)

The Austrian company Repotec is a technology provider and its

integrated technology platform includes two interconnected fluid

bed reactors. The first is a steam blown gasifier and the second a combustor which burns the resulting

coke byproduct (Repotec, 2013). The syngas is fed to a gas IC engine and the flue gas heats an Organic


July 2014 132

Rankine Cycle (ORC) turbine that generates additional power. The system is designed for maximum

power generation and it has been operating at the Guessing facility in Austria since 2002 (8 MWth, 2

MWel). Another plant, which is under commissioning at the time of writing for Senden, Germany (14

MWth, 5 MWel) will provide power for 21,000 inhabitants (Repotec, 2013).

Rentech (www.rentech.com) has purchased the rights to the dual fluid

bed gasification process known as “Silvagas”. The process was first

developed at Battelle’s Columbus Laboratories in the US (thus also

known as the “Battelle Process”) and the rights were sold in 1992 to

FERCO Enterprises and in 2001 to Biomass Gas & Electric. This dual fluid bed technology (shown in the

schematic below) is comprised of two fluid bed reactors one of which is the gasifier (which is steam

assisted and operates in the absence of oxygen) and the other the combustor (which operates in

conditions of excess oxygen). The gasifier sends sand and char particles to the combustor which are

entrained in the flowing gas (particles are separated from product syngas with the use of a cyclone). The

combustor then burns the char and returns clean hot sand to the gasifier (isolated from the flue gas with

the use of another cyclone).

The advantage of this configuration is that the gasification can take place in the absence of oxygen

inputs, thus minimizing oxidation reactions on the biomass feed. The oxidation step discussed in Section

http://www.rentech.com/


July 2014 133

3.7.1 is necessary to produce the heat to drive the final syngas producing reactions. However, in the

Silvagas process this heat is indirectly supplied by the combustor reactor which heats the fluidized sand

by combusting the char residues coming from the gasifier (E4tech, 2009). Rentech added syngas clean-

up and Fischer-Tropsch fuel synthesis components to the technology and recently entered the drop-in

biofuels commercialization arena. However, in early 2013, Rentech’s plans for a commercial drop-in

biofuel facility in Natchez, MS, USA were shelved and its R&D facilities in Colorado were also closed.

Other opportunities with more near term returns were the company’s reason for these developments

(Lane, 2013b).

3.10.5 Entrained Flow Gasification facilities

Choren, a German company established in the early 1990’s,

developed the Carbo-V process, which is an entrained flow

biomass gasification technology comprising three process

steps (see Figure 4-6). The first step involves slow pyrolysis of the raw biomass at 400-500 °C, producing

mainly gases and char. The gases are fed into a second high temperature (1200-1500 °C) chamber where

the tars are cracked or converted to permanent carbonaceous gases. The char from the first step is fed

together with the cracked gases from the second chamber into the third entrained flow reactor which

operates at 700-900 °C. The main advantage of this multistage approach is that the final EF gasifier

operates with an optimized and pulverized solid feed entrained in a tar-free carbonaceous gas. The

process has been reported to perform well with thermal efficiencies of 91% and cold gas efficiencies of

81%. In 1998 Choren built a 1 MWth pilot plant in Germany, Freiberg which was operated for thousands

of hours, including the production of liquid fuels using Shell’s gas-to-liquids process. Scale-up from 1 to

45 MWth was begun in 2003 but it was never completed. Reported reasons included the long time it

took to troubleshoot the scale-up of each process step and the resulting loss of investors’ patience.

These developments culminated in Choren filing for bankruptcy in 2011 and selling the Carbo-V process

to Linde engineering in 2012 (E4tech, 2009; Linde Engineering, 2012).

Choren’s Carbo-V is also the gasification technology of choice for the biofuel facility

proposed by Finnish company Forest-BtL Oy, a subsidiary of Vapo Group. The 480

The Silvagas process (E4tech,2009))


July 2014 134

MWth facility has just received 170 million Euro worth of NER300 funding and is due begin operations in

2016-2017 (Landalv, 2013).

Figure 3-14: The Choren gasification/FT process Source: (E4tech, 2009)

Range fuels, a Khoshla Ventures company, is another entrained flow

gasification technology which has also encountered scale up

challenges culminating in the closure of the company in late 2011.

The process, originally named the “Klepper Pyrolytic Steam

Reforming Gasifier”, after the surname of its US inventor Robert Klepper, was marketed under the name

“K2”. Two separate reactors were used, the “devolatilization” reactor (low temperature gasification) and

the reforming reactor (gasification). In the first reactor the biomass was heated to 230 °C where the

most reactive proportion of the biomass (possibly most of the hemicellulose) was removed

(devolatilized) in the form of flue gases. The feed temperature continued to rise until it was entrained

with an 815 °C steam stream and fed into the main gasifier. The reported thermal efficiency of the

process was around 75%. Unique features that contributed to Range Fuels’ efficiency and compact

design included the recycling of the steam and part of the syngas to be used as carrier gases in the next


July 2014 135

gasification cycle and the incorporation of the cyclone filters and water condensers inside the main

reactor. In 2009 Range Fuels began building a 150 MLPY (40 MGPY) facility at its Soperton site in

Georgia, US. However, it encountered severe tar accumulation and scale up challenges. The size of the

facility was subsequently revised down to a 75 MLPY (20 MGPY) facility and then to a 15 MLPY (4 MGPY)

methanol plant instead of the promised ethanol facility. The project received more than $160 million in

investor funding, plus $162.25 million in government commitments. Aside from this loan guarantee the

government provided an additional $76 million DOE grant in 2007 and a $6.25 million state grant. In

total, the DOE released $43.6 million of the project funds before suspending payments and terminating

the agreement with Range in August 2011. In an attempt to salvage part of the invested funds, Range’s

assets were transferred by its investor group to LanzaTech (a syngas fermentation company that is

discussed later) at the end of 2011 (Lane, 2011b).

The KIT bioliq® process was developed in a project carried out at the

Karlsruhe Institute of Technology and supported by a collaboration

between Lurgi AG and Future Energy GmbH. Lurgi is now part or the Air

Liquide Group (since 2007) and Future Energy part of the Siemens Power

Generation Group (since 2006). The KIT bioliq® process has been discussed in the pyrolysis section of the

report. The main principle is producing energy dense biomass slurries in separate auger-pyrolizers and

then feeding these “bioliq SynCrude” slurries into a central entrained flow gasification facility. The

resulting syngas is purified using Lurgi’s Rectisol and Purasol processes. Testing has been carried out

since 2006 using a 2MWth bio-slurry pilot plant. In 2007 a second phase was approved and funded in

order to build a 5MWth facility that included an 85 bar gasifier. The second phase completion was

announced in February 2013 with successful operation and the ability to produce very high quality,

almost “tar-free”, biomass syngas (KIT, 2013).

Chemrec, a Sweden-based gasification company, has specialized in

technology development for black liquor gasification (see box 3-1

below). A 3 MWth pilot plant has been installed next to the Smurfit

Kappa pulp mill in Pitea, Sweden since 2005. This pioneer facility has

collaborated with Haldor Topsoe to convert its syngas to Dimethyl Ether (DME). They reached 4 t/day of

DME production in H1 2013. Volvo has tested the DME in truck engines with encouraging results. The

Chemrec facility is now part of the EU’s 28 million Euro 7th framework programme BIODME project


July 2014 136

which includes various high profile collaborators such as Haldor Topsoe, Total, Preem AB and Volvo

(BioDME, 2013). The company, COWI (which is the commercialization agent for the Chemrec technology

as it has evolved at the Pitea pulp mill) has carried out a feasibility study to produce 55 000 tons of

methanol for New Page in Michigan, USA. The plant will use 30% of the black liquor flow from the pulp

mill (COWI, 2013). COWI has also performed a feasability study and cost estimate for a full scale biofuel

production plant, including a black liquor gasification unit, to be located at Domsjö Fabriker, Sweden.

Although neither methanol nor DME can be considered drop-in biofuels, they are direct precursors to

gasoline and their conversion to drop-in fuels is technologically mature and proven in the chemicals

industry. Lulea University of technology’s holdings company has recently (February 2013) acquired

Chemrec and its Pitea pilot plant. LTU Green Fuels will be the name of the new company (RISI, 2013).

Box 3-1: Black liquor gasification: Improving thermal efficiency and value streams for pulp mills

Kraft mills produce a highly carbonaceous pulping process effluent known as “black liquor”. This

liquid stream contains process chemicals as well as half of the carbon that was present in the

starting biomass, mainly in the form of lignin (lignin gives the liquid its dark brown colour hence the

name “black”) (IEA Bioenergy, 2007). In Kraft mills, black liquors are concentrated and combusted in

large recovery kilns where the inorganic chemicals are recycled and the heat and power generated

are used to feed the pulping process. Due to recent sustainability and bio-economy trends, pulp mills

have been considering replacing their recovery kilns with entrained flow gasifiers in order to improve

their energy efficiency and provide them with the flexibility of utilizing their syngas to diversify their

product range into liquid fuels and chemicals (via for example Fischer-Tropsch synthesis). Over the

last couple of years Canada and the US have used subsidies (e.g. Black liquor subsidy) to encourage

their pulp and paper mills to enhance the utilization of their black liquors and improve their business

competitiveness. Although various pilot plants have been installed in pulp mills in Sweden and the

USA, mostly using Chemrec technology (the largest being the 50 t/day black liquor gasification

demonstration facility at Weyerhauser’s New Bern pulp mill in N. Carolina, USA, (Babu, 2005)), there

appear to be no commercial sized black liquor gasifiers operating at this time. Part of the reason is

that only large scale facilities that accumulate black liquor equivalent to more than 1000 t of dry

solids daily are practically able to consider the conversion to gasification (IEA Bioenergy, 2007).


July 2014 137

3.10.5.1 Biomass Co-gasification with coal

Entrained flow gasifiers are widely used for coal gasification and thus have been used in many biomass

co-gasification applications. Co-gasification allows for mitigating some of the biomass gasification issues.

Examples of currently operating co-gasification plants include the:

230 MWe and 170 MWth, The Vaasa, Finland, facility, cofires up to 40% forest residue in an

existing pulverized coal utility boiler. The facility has been in successful operation since

December 2012 (Simet, 2013).

50 MWe & 90 MWth, Lahti Energia (MSW feedstock) facility in Finland, Metso technology

(Breitholtz, 2013; Simet, 2013)

140 MWth Vasikuodon Voima facility in Finland, Metso gasification technology (Breitholtz, 2013)

80 MWe Essent’s Amer Power Plant in the Netherlands (Essent, 2013)

3.10.6 Plasma gasifiers

Although plasma gasifiers have a similar structure to entrained flow reactors they operate using plasma

torches at extremely high temperatures of 1500 °C – 5000 °C (atmospheric pressure). Although they are

supposedly costly to operate, they are able to produce a very high quality syngas which needs little

further cleanup (E4tech, 2009). In this process, the biomass is converted to syngas and the inorganic ash

is vitrified to inert slag. Although this technology is being developed by various start-up companies

including California-based Solena and US-based InEnTech the technology is still at the development

stage. InEnTech has been selected as a gasification technology provider for the California-based Fulcrum

bioenergy (E4tech, 2009).

3.10.7 The IEA Bioenergy Task 33 gasification facilities database

For an extensive list of global demonstration and commercial gasification facilities, the reader is referred

to IEA Bioenergy’s Task 33 (Gasification) online database. This online tool (see Figure 3-15) features an

interactive world map with the gasification facilities at demonstration or operation stage. According to

this interactive map, most of the biomass gasification activities are concentrated in Germany, Austria

and the Scandinavian countries.

http://www.ieatask33.org/content/thermal_gasification_facilities


July 2014 138

Figure 3-15: The IEA Bioenergy Task 33 gasification facility database

Biomass gasification has developed substantially over the last couple of decades as indicated by the

more than 500 MW of operating biomass gasification CHP facilities around the world. The biogas-to-

distillates end of the process will also benefit from industrial “know-how” based on the operation of

Sasol coal-to-liquids facilities. However, the upgrading of syngas to transport fuels remains a significant

challenge primarily due to high capital costs. These costs are due to the high temperature and high

complexity of the gasification platforms which typically include various heat-cool and compression

cycles. To try to mediate these high capital costs, it has been suggested that a minimum scale of 2000

tpd size facility will be needed. However, this can be considered to be relatively small scale when

compared to current, commercial fossil based FT-distillates facilities which are at least 50 times larger. It

can be anticipated that FT synthesis facilities which gasify biomass will be equally, if not more,

challenging to operate at scale than are their fossil counterparts (CtL and GtL). However, a major

advantage of gasification is that the perceived technology risk will likely be lower than that of the other

biofuel platforms because, from a chemical standpoint, syngas is sufficiently similar to commercially

upgraded fossil syngas. This perception might be challenged, though, as biomass syngas is generally of

inferior quality to coal-derived syngas and especially natural-gas-derived syngas. For example, the tar

content in biomass syngas (gasification and tar conversion technology dependent, up to thousands of

ppm) can be much higher than in coal-derived syngas and orders of magnitude higher than can be

accommodated in FT synthesis (about 5 ppm) to biofuels. Tar accumulation, capital costs and scale up

http://www.ieatask33.org/content/thermal_gasification_facilities


July 2014 139

challenges have been suggested to be the main reasons behind the recent failure of the three

commercial size start-up, biomass-to-distillates gasification facilities of NSE biofuels, Range Fuels and

Choren (Lane, 2011b; Rapier, 2011; Lane, 2012; Hayes, 2013; EBTP, 2013). Techno-economic studies

have shown that the manufactured selling price (MFSP) of gasification biofuels can be between $31-

$36/GJ while the capital costs (TPI) required to build a pioneer commercial facility can be in the vicinity

of $1 billion for a drop-in gasification plant processing 2000 tpd of biomass (dry basis). However, there is

still significant potential for gasification technologies as new systems and technologies are developed.

From a systems point of view, the “hub and spoke” model of biomass logistics may provide an

opportunity for a gasification facility to be built at a very large scale where pyrolysis will provide the

densified biomass feed and oil refineries will be used for the last upgrading steps of the FTS process.

From a technology point of view, researchers and companies, (such as Velocys which claim efficient FTS

of biomass at small scale (<500 tpd)) continue to make significant progress in all aspects of the process.

For example, cleaning technology (e.g. KIT claiming production of “tar free” syngas from entrained flow

gasifiers) and an increased interest in dimethyl-ether as an alternative fuel to diesel (e.g. the “BioDME”

project in Europe) have already made an impact on the technical end economic feasibility of this

approach to drop-in biofuel production (E4tech, 2009; Rapier, 2011; Linde Engineering, 2012).


July 2014 140

CHAPTER 4: THE BIOCHEMICAL PLATFORM

Biomass conversion through biological means, such as by enzymatic hydrolysis followed by biological

sugar conversion, has been termed the “biochemical” approach to producing drop-in biofuels. In the

same way that glucose is fermented to ethanol for conventional (sometimes termed first generation)

biofuel production, these “advanced” biocatalytic processes convert sugars to less oxygenated, more

energy-dense molecules such as longer chain alcohols (butanol, butanediol) and higher molecular

weight compounds such as isoprenoids and fatty acids. As shown in Figure 4-1, there are numerous

biological pathways, feedstocks, and microorganisms that have been proposed for the production of

drop-in biofuels and their intermediates. Microorganisms such as cyanobacteria and algae can directly

capture CO2 from the atmosphere and convert it to ethanol or lipids. Alternatively, bacteria, yeasts and

heterotrophic algae can utilise sugars derived from sugar cane, sugar beet, starch and other energy

storage polysaccharides, or from the hydrolysis of the cellulose and hemicellulose carbohydrates in

cellulosic biomass. Other bacteria can utilise hydrogen and carbon monoxide in syngas. Although the

intermediates produced can be used for diesel or gasoline production, generally more oxygenated

higher alcohols are better suited for gasoline production and higher molecular weight, longer chain

more saturated lipids and isoprenoids are better suited for energy dense diesel and jet fuel fuels.

Cyanobacteria, yeasts and bacteria can be selected or engineered to produce either higher alcohols or

isoprenoids and lipids and the metabolic pathways from one organism can be heterologously expressed

in another. However, after biological production, regardless of which biosynthesis route is used, some

form of hydroprocessing of the produced intermediate is typically required before blending with

conventional petroleum fuels. As described earlier, the more oxygenated and unsaturated the

intermediate, the more hydrogen is required for it to be upgraded to a fungible drop-in functional

equivalent to diesel, jet or gasoline blendstock.

From a commercialization point of view, one of the most attractive characteristics of these technologies

is the potential to “piggy-back” onto existing ethanol facilities by switching the microorganism being

used in the microbe-to-ethanol process to obtain a “microbe-to-drop-in fuel or fuel intermediate”

process. However, high process efficiencies must also be achieved and the industrial robustness of the

process at commercial scale must be proven. These and other aspects of the various emerging biological

pathways to drop-in biofuels are still evolving.


July 2014 141

Figure 4-1: Schematic overview of the microorganism-to-drop-in-intermediate or biofuel derived for various renewable feedstocks Source: adapted from (Weber et al., 2010)

4.1 Ethanol

Ethanol, the world’s most widely used biofuel, is currently biochemically produced from sugars or starch

by varieties of the yeast Saccharomyces cerevisiae. Although ethanol is not a drop-in biofuel it is the

benchmark process against which alternative biological conversion pathways should be compared. A

typical bioethanol process entails three steps:

CO2 and

sunlight Plant biomass

Free sugars, starches and pectins Lignocellulosics

Syngas C6 sugars C5 sugars

Algae

Cyanobacteria

Synechocystis

Synechococcus

Bacteria

Zymomonas

Clostridium

E. coli

Yeasts

Saccharomyces

Pichia

Lipids, isoprenoid derivatives Ethanol, Butanol and higher alcohols

Gasoline Diesel / Jet

FEEDSTOCK

Micro

organ

isms

PRODUCT


July 2014 142

Either direct utilization of the sugar or the hydrolysis of sucrose or starch (or cellulose and

hemicellulose in the case of non-food lignocellulosic feedstocks) to monosaccharides such as

glucose.

The metabolic conversion (anaerobic fermentation) of glucose (or other monosaccharides) to

ethanol.

The recovery and concentration of ethanol from the fermentation broth (beer).

The “ethanologenic” microbe itself typically contains the invertase enzyme needed to break

disaccharide sucrose (derived from cane or sugar beet) down to its hexose monomers (fructose and

glucose). For starch hydrolysis, externally sourced “amylase” enzymes are usually used and this is the

dominant route practiced in the US corn ethanol industry (which, as discussed in Chapter 1, is currently

the world’s largest ethanol production model; sucrose from sugarcane to produce ethanol is the second

largest). A more complex process is required to hydrolyse lignocellulosic carbohydrates to fermentable

hexose and pentose monosaccharides. There are many approaches, but all are technically and

economically challenging, especially as in many lignocellulosic feedstocks as much as 30 wt% of total

sugars are pentoses (primarily xylose) that are not readily converted to ethanol by wildtype

microorganisms. A detailed description of a typical “cellulosic ethanol” process that involves

pretreatment, enzymatic hydrolysis, fermentation, product recovery and residual solids utilization, etc.,

is beyond the scope of this report and the reader is referred to other publications that more extensively

review options and progress in cellulosic ethanol development (Humbird et al., 2011; Saddler et al.,

2012).

The main metabolic pathway used by yeasts in ethanol fermentation is the Embden–Meyerhof–Parnas

or EMP pathway. The pathway involves an initial glycolysis step, where one molecule of glucose is

metabolized, and two molecules of pyruvate are produced. Each pyruvate is then metabolized to one

molecule of ethanol (Bai et al., 2008). The overall glucose-to-ethanol reaction is stoichiometrically

represented by the equation below:

C6H12O6 + 2ADP + 2 Pi → 2C2H5OH + 2 CO2 + 2 H2O + 2 ATP

This equation shows that four of the six carbons in the hexose sugar feedstock end up in the ethanol

product and two are oxidized to CO2 (on a mass basis, 49% is converted to ethanol and 51% to CO2). The

carbon oxidation to CO2 displaces carbon from ethanol production but gives rise to usable energy in the

form of two ATP molecules while it also generates the reducing power necessary for the chemical


July 2014 143

reduction (“deoxygenation”) of the sugar to the ethanol molecule. The stoichiometric conversion yield in

the above equation is theoretical, meaning lower conversion yields are achieved in practice. This is

because some of the carbon is used as structural material for actively growing microbial cells and the

metabolic efficiency of this conversion may be compromised by various mechanisms such as feedback

inhibition, toxic chemicals, biological contaminants, osmotic pressure stress or inadequate nutrition

(Sanchez and Cardona, 2008). Yeasts are the “workhorse” microorganism used in the current ethanol

industry, as they have a long history of use and routinely achieve more than 90% of the above

theoretical conversion in non-sterile industrial settings (Ingledew, 2008). Traditionally, alcohol plants

produce a “beer” of 5% – 10% v/v ethanol/water concentration. However, more recent plants are able

to produce beer of up to 20% volume ethanol (by carefully monitoring conditions and dosing yeast

nutrients) (Ingledew, 2008; Walker & Brew, 2011). This performance is unmatched by other

fermentative microorganisms and it is attributed to a number of unique characteristics exhibited by

Saccharomyces cerevisiae’s ability to convert glucose.

These yeasts have evolved to produce large amounts of ethanol and to tolerate high

concentrations of ethanol. In fact, 20 vol% ethanol concentrations have been reported in

industrial settings without the occurrence of significant feedback inhibition.

These yeasts are relatively resistant to toxic inhibitors such as organic acids, aldehydes,

inorganic salts, etc.

Their cell growth consumes little carbon and it is directly associated with vigorous ethanol

production. This contrasts with many fermentative microorganisms which only produce

alcohols under stress and as secondary metabolites.

They ferment sugars at low pH values, which minimizes contamination.

They are relatively fast growing microorganisms which adapt readily to changing fermentation

conditions.

They are highly selective in producing ethanol and CO2 and only minute amounts of carbon are

diverted to other fermentation metabolites such as organic acids, esters and aldehydes

(Ingledew, 2008; Walker & Brew, 2011)


July 2014 144

Alternative microorganisms are also being evaluated and developed for ethanol production such as

Zymomonas mobilis, which is a rapidly fermenting ethanologenic bacterium, and thermophilic bacteria

that can operate at the higher temperatures typically used for enzymatic hydrolysis of cellulose.

Zymomonas mobilis is a highly productive and selective ethanologen bacterium that has been studied

since the 1970s. This bacterium is used to ferment sugar-rich plant saps such as obtained from agave

cacti to produce the traditional pulque drink of Mexico (Swings and Deley, 1977). It is an anaerobic,

gram negative bacterium. Instead of the EMP pathway used in yeast, it uses the Entner-Doudoroff

pathway, producing only one ATP (instead of two in the EMP pathway) per molecule of glucose

metabolised. The lower ATP production means that less ATP is accumulated intracellularly and less

carbon is funneled into growth of cell mass, leaving more available for ethanol production. It also means

that Z. mobilis strains must achieve higher specific rates of glucose utilization than yeast. Z. mobilis

tolerates up to 120 g/L of ethanol and its specific productivity is 2.5-5 times higher than traditional

baker’s yeast (S. cerevisiae) and is able to produce ethanol at yields approaching 97% of theoretical (Bai

et al., 2008). While Z. mobilis ferments glucose efficiently, fructose and sucrose and mixtures of biomass

sugars are more challenging feedstocks. Genetically engineered Z. mobilis strains have been developed

that are able to ferment pentoses as well as hexoses. An engineered Z. mobilis strain is currently being

assessed at DuPont Cellulosic Ethanol’s (DCE) 250 KGY cellulosic ethanol facility in Vonore, Tennessee

(USA). A possible limitation on the commercial use of Z. mobilis for fermentation of corn starch is that

the process residual Distillers Dry Grains and Solubles (DDGS) would need to be proven to be acceptable

for animal feed. Sales of DDGS present an important co-product revenue stream and can account for up

to 20% of a corn ethanol mill’s total revenue (Bai et al., 2008, Walker & Brew, 2011).

Thermophilic bacteria such as Thermoanaerobacterium saccharolyticum are able to tolerate higher

temperatures than 30-35◦C which is the optimum temperature range for most mesophilic yeast. An

ablity to operate at higher temperatures can have numerous process advantages, especially for enzyme

based cellulosic ethanol production. In particular, enzymatic hydrolysis of cellulose is typically

performed by cellulolytic enzymes that exhibit temperature optima of 45-50 °C or even higher. The

discrepancy in temperature optima between the sequential steps of enzymatic hydrolysis and

fermentation creates a need to cool the process medium prior to fermentation. In contrast,

thermophilic ethanologens are able to operate at the same temperature as the hydrolytic enzymes,

circumventing the need for cooling. Lower feedback inhibition on the enzymes is an additional

advantage of this simultaneous saccharification and fermentation approach (SSF) as glucose is


July 2014 145

simultaneously produced and converted in the same process step. However, to date, the performance

of thermophilic bacteria that have been investigated for this application is below that of S. cerevisiae

and Z. mobilis, and their ethanol tolerance is also significantly lower (cf 2%-5% vol for thermophilic

bacteria vs up to 20% vol or higher for industrial strains of S.cerevisiae) (Weber et al., 2010).

Ethanol has a high vapour pressure and can readily be recovered and concentrated by distillation of the

aqueous fermentation beer. However, ethanol forms an azeotrope with water such that distilled ethanol

still contains 3 wt% or more water. This remaining water can be removed (if it needs to be) through

molecular sieves or other dehydration technologies (e.g., pervaporation or ternary distillation), which

are widely practiced at industrial scales. Ethanol recovery and purification are relatively well understood

processes compared to recovery of other potential biofuels such as butanol and higher molecular weight

compounds which have lower vapour pressures and are produced at lower maximum titers.

Despite ethanol’s established industrial processes and biofuel markets, the ethanol molecule is too

oxygenated to be a drop-in biofuel. So called “advanced biological conversion” pathways can produce

less oxygenated (more reduced, higher Heff/C) molecules such as higher alcohols (e.g., butanol), fatty

acids (e.g., palmitic acid) and bio- hydrocarbons (e.g., farnesene). Although these latter molecules are

better suited as intermediates for drop-in biofuel production, their production will still require the

resolution of significant operational and scale up challenges compared to sugar, starch or cellulosic

ethanol production processes.

4.2 Biological conversion to drop-in biofuels

As described earlier, there are four main metabolic pathways that have been proposed for the

conversion of glucose to the chemically reduced molecules that could constitute drop-in biofuels. The

four pathways are shown in Figure 4-2 and they all begin with the oxidation of glucose to two pyruvate

molecules (glycolysis). From there the pathways differ as the pyruvate is converted to various

intermediates such as Acetyl CoA or acetaldehyde. The pathways that lead to butanol and alcohol are

anaerobic whereas the pathways that lead to more saturated longer chain molecules such as

isoprenoids or fatty acids are aerobic (Jin et al., 2011). Each pathway plays a different role in the

production of drop-in biofuels in terms of productivity and suitability for drop-in biofuel production.


July 2014 146

Figure 4-2: Simplified schematic of major metabolic pathways relevant to drop-in biofuel production

4.2.1 The butanol pathway

Although butanol is not directly a drop-in fuel, it is considered to be a more favorable drop-in biofuel

intermediate than ethanol because it is less oxygenated (lower O/C ration and higher Ceff/H). Butanol is a

four carbon primary alcohol with a volumetric energy content of 28 MJ/L which is higher than ethanol’s

19 MJ/L and closer to gasoline’s 36 MJ/L (Pfromm et al., 2010). Other favourable characteristics of

butanol when compared to ethanol are its lower volatility and less hydroscopic and corrosive properties.

The octane number of normal butanol (n-butanol) is about 87 while branched chain butanols

(isobutanol) have an even higher octane number (Nexant, 2012). Although the butanol molecule is not a

1 (aerobic)

Glucose

Pyruvate Pyrophosphate

intermediates

DMAPP, IPP

Acetyl CoA

Acetoacetyl-CoA

Acetaldehyde

Ethanol Fatty

acid

Butanol

Isopropenol

1. Isoprenoid pathway (aerobic)

2. ABE pathway (anaerobic)

3. Fatty acid pathway (aerobic)

4. Ethanol pathway (anaerobic)

+ O2 decarboxylation

- O2 reduction

4 (anaerobic)


July 2014 147

true drop-in fuel component, it can be directly blended with gasoline at high volumetric ratios (Jin et al.,

2011).

Butanol (n-butanol) is a commodity chemical with a current 3 Mt/year global market valued at USD $7.5

b/year based on a spot price of USD $2500/t) (Indexmundi, 2013). It is used in diverse applications such

as coatings, plasticizers and solvents and is currently produced from petroleum using the oxo-synthesis

(hydroformylation) process (Machado, 2010). The raw material, propylene, reacts with carbon monoxide

and hydrogen over catalysts (Co, Ru and Rh) to yield aldehyde mixtures. These aldehydes are then

hydrogenated (catalytically) to produce the final product which is a mixture of butanol isomers. The

BASF and Dow Chemical companies are currently the world’s leading butanol manufacturers and jointly

account for about 35% of global production (Machado, 2010).

From 1914 up until WWII, butanol was primarily produced via fermentation using the bacterium

Clostridum acetobutylicum (Qureshi & Blaschek, 1999). This microorganism is able to convert a wide

range of sugars (including pentoses) to acetone, butanol and ethanol (known as the “ABE” fermentation

process) at typical ratios of about 3:6:1 by weight. Initially coatings and paints were the main uses of

these solvents, although during WWII butanol was also used as a fuel for British Royal Air Force planes

and acetone was used to manufacture cordite, a smokeless type of gunpowder (Machado, 2010). With

the development of cheap petroleum in the 1950s, most of the world’s biological route butanol (ABE)

fermentation facilities closed (other than some facilities in Russia and South Africa which continued to

operate up to the 1980s) in favour of manufacturing butanol from petroleum using the chemical process

(Lee et al., 2008).

The fermentation of sugars to n-butanol has been advocated as a potential platform to produce drop-in

biofuel intermediates. The equation below describes the theoretical stoichiometric conversion of

glucose to butanol via the ABE pathway.

C6H12O6 + 2ADP + 2 Pi → C4H9OH + 2 CO2 + H2O + 2 ATP + CoA

Although the theoretical carbon conversion is the same as for the EMP pathway used by yeasts (2/3 of

the sugar carbon ends up in the alcohol product), Clostridium sp. have, so far, been challenged to reach

this theoretical yield in industrial practice. Although the theoretical butanol yield is 41% by mass, a mass

yield of only about 15% is typically achieved (Jang et al., 2012; Lee et al., 2008). Clostridium species also

tend to be slower-growing and, while they can exhibit a wide sugar substrate range, they are susceptible


July 2014 148

to solvent toxicity by butanol (product feedback inhibition). Solvent toxicity is arguably the biggest

challenge of the ABE metabolic pathway and industrial bio-based butanol production process as cellular

inhibition becomes severe at butanol concentrations higher than 10-20 g/L (Jin et al., 2011). This is

primarily a result of the hydrophobic nature of the butanol, which enables it to penetrate into the lipid

bilayer of cell membranes and disrupt cellular function. It is also hydrophilic enough to remain in

aqueous solution up to concentrations of 70-80 g/L, although at concentrations above this it phase

separates and creates a butanol-rich upper “oil-like” layer in aqueous systems (Jin et al., 2011). Genetic

engineering efforts have already achieved a more-than-doubling of butanol titers, from 8 g/L in early

wild type strains to about 20 g/L (2% w/v) in improved stains (Peralta-Yahya & Keasling, 2010; Weber et

al., 2010). However, this is still a low titer and almost an order of magnitude lower than what can be

achieved in the ethanol industry today (16% w/v or 20% vol) (Walker & Brew, 2011).

The lipophilic nature of butanol makes it much more toxic to microbial systems than ethanol. As it

accumulates intracellularly, it disrupts phospholipid components of the cell membrane causing an

imbalance in cellular metabolism by disturbing the selective flow of metabolites, nutrients and toxics in

and out of the cell (Jin et al., 2011). Other challenges to using Clostridium sp. in industrial settings are

their relatively slow growing nature and susceptibility to bacteriophage attack and cell degeneration, as

well as a requirement to maintain strictly anaerobic conditions (Weber et al., 2010).

To try to circumvent these drawbacks, some technology developers have heterologously expressed the

ABE pathway in industrially robust yeasts (Jin et al., 2011; Peralta-Yahya & Keasling, 2010). Although

these genetically modified microorganisms are expected to be more amenable to large scale industrial

processing conditions, the solvent toxicity issues of ABE pathway products will continue to be a

challenge to achieving robust production performance. While n-butanol is the isomer that is naturally

made by the ABE pathway, some companies such as Gevo and Butamax have engineered pathways in

yeast cells to produce isobutanol. This isomer has proven to be easier to produce in yeast and it is also

more valuable as a biofuel feedstock as a result of having a branched carbon chain. Although little is

known about the scale up performance of these isobutanol producing pathways, Gevo has obtained a

fivefold higher productivity than they obtained for n-butanol (Nexant, 2012).

Another strategy to overcome solvent toxicity of butanol (or ABE pathway products) is to design

production systems that enable continuous product removal. In situ removal strategies include liquid-

liquid extractions using co-solvents, gas stripping and the use of membrane based separations. Ezeji and


July 2014 149

co-workers are working on reactor designs that are able to perform continuous butanol product removal

(Ezeji et al., 2010). Ranjan and Moholkar (2012) recently reviewed solvent extraction strategies and

concluded that the most promising technologies are gas stripping and pervaporation. However, both

processes have certain drawbacks. In particular, the butanol recovered by gas stripping contains

relatively high levels of water and further azeotropic distillation is required to recover butanol of

sufficient purity (Liu et al., 2004) while pervaporation requires specialized membranes that are, to date,

easily fouled by fermentation broth components(Lee et al., 2008). Although novel techniques such as a

hybrid vapor stripping/vapor permeation process have been proposed (Vane et al., 2013), such

techniques can consume more energy than contained in the butanol molecule itself. As the efficiency of

these recovery techniques is highly sensitive to the butanol titer (concentration), improvements in

process titers will enhance both butanol fermentation performance and product recovery efficiency.

4.2.2 The fatty acid pathway

Fatty acids are precursors to some of the best known storage lipids, namely triacyl glycerides (TAGs).

TAGs are formed by three long alkyl chains (C12 to C22) connected to a glycerol head via ester bonds

(see Figure 2-1). As discussed in Chapter 2, TAGs are found in the vegetable oils of oilseed plants (such

as palm, jatropha, rapeseed/canola and soy) and can be used to produce diesel and jet fuel blendstocks.

Many diverse microbes including green algae, bacteria and fungi can also produce high levels of TAGs

through specialized metabolic pathways (Peralta-Yahya & Keasling, 2010). As shown in the simplified

pathway for palmitate fatty acid production (Figure 4-3), fatty acid biosynthesis involves chain

lengthening thiol-esters produced by low molecular weight acyl carrier proteins (ACP) (Jovanovic Tews

et al. 2010). Up to the formation of butyryl-ACP, this pathway resembles the butanol production

pathway. In order to elongate the 4-carbon butyryl chain to a 16-carbon palmitate chain, 6 elongation

cycles are required (Jovanovic Tews et al. 2010). With each elongation cycle, two carbon units are

sequentially condensed onto the growing fatty acyl chain. Each elongation cycle expends the reducing

power of two NAPDH molecules. Replenishing this reducing power typically requires a portion of

feedstock carbon to be oxidized to CO2 and H2O via aerobic respiration (Jovanovic Tews et al. 2010). The

NADPH generation process takes place within the pentose phosphate pathway (PPP), as shown in Figure

4-3. The reducing power of NADPH essentially provides a means of hydrogen insertion. In fact, inserting

molecular hydrogen via alternative means has been shown to increase yields (and decrease the need for

carbon-consuming NADPH generation) in metabolic processes involving chemical reduction of

carbohydrates (Yerushalmi et al., 1985). The longer and more saturated the fatty acid chain, the more


July 2014 150

elongation cycles and greater reducing power required. This reducing power has to come from either

alternative hydrogen inputs or from oxidation of feedstock carbon which sacrifices yield.

Figure 4-3: Biosynthesis of fatty acid, palmitate Source: adapted from Jovanovic-Tews et al. (2010) Cyanobacteria (also known as microalgae) can also be used to accumulate high levels of TAGs via the

fatty acid pathway. Autotrophic cyanobacteria use photosynthesis to convert atmospheric CO2 to

glyceraldehyde-3-phosphate, which in turn forms pyruvate through the same glycolytic process used to

metabolize glucose in heterotrophic microbes. Thus, these photosynthetic microbes are able to utilise

Reducing power from

oxidizing carbon at the

PPP cycle

Acetoacetyl-ACP

HCO3-

Glucose

(C6)

Pyruvate (C3)

Acetyl-

ACP

(C2)

Malonyl-ACP

(C3)

Malonyl-ACP

ACP + HCO3-

2 x NADP+

2 x NADPH

Palmitate (C18)

Butyryl-ACP (C4)

ACP

Elo

nga

tio

n c

ycle

s:

Re

pe

at 7

x


July 2014 151

the most oxidized form of carbon (CO2) and convert it to highly-reduced energy-dense triacyl glyceride

molecules. As cyanobacteria do not require soil or arable land to grow, it has been suggested that, on a

per unit land basis, they are potentially higher yielding than any terrestrial oilseed crop. However, there

are several technical challenges to achieving large scale economical cultivation and harvest of algae, as

reviewed in a recent Task 39 report (Darzins et. al, 2010). Despite these challenges, considerable

interest remains in advancing algal-based route as indicated by recent funding awards such as:

The European Commission’s 7th framework programme which supported 3 algae contracts for a

total of €19.2 million under the “Algae Cluster” initiative. “Biofat” was supported with €7.1

million, “ALL GAS” with €7.1 million and “Intesusal” with €5 million;

The US DOE’s $50 million USD award to the US’s National Alliance for Advanced Biofuels and

Bioproducts (NAABB); and

The $600 million USD research alliance between ExxonMobil and Craig Venter’s Synthetic

Genomics company to develop high performing algae for biofuels production.

4.2.3 The isoprenoid pathway

Isoprenoids are members of the lipid class of molecules and they can also be used as intermediates for

diesel and jet blendstock biofuel production. As their carbon chains are more branched than fatty acids

they exhibit lower viscosities, lower boiling points and higher octane numbers and are thus better suited

for conversion to transport fuels than straight chain paraffinic molecules (e.g., dodecane) (Renninger et

al. 2010). The main metabolite produced from the isoprenoid pathway is isopentenyl diphosphate (IPP)

(Figure 4-4) and its isomer, dimethyl allyl diphosphate (DMAPP). These 5-carbon intermediates are

biologically condensed by phosphatase enzymes to form longer chain and branched alcohols or alkenes

such as the 15-carbon alkene farnesene (Jovanovic Tews, 2010). Most efforts to date on exploiting the

isoprenoid pathway have focussed on optimizing production of the 15-carbon amorphadiene, a key

precursor to the synthesis of various pharmaceuticals (Peralta-Yahya & Keasling, 2010). The biggest

challenge in these processes is maximizing yield of C5 IPP and DMAPP precursors. In nature, only trace

amounts of these compounds are made by prokaryotes such as E. coli via pyruvate using the deoxy-

xylulose (DXP) pathway. This pathway is the only one in Figure 4-2 that does not involve the formation

of Acetyl-CoA. However, Martin et al. (2003) have increased yields of IPP and DMAPP through formation

of Acetyl-CoA by heterologously expressing the mevalonate pathway of Saccharomyces cerevisiae in E.

coli. Expressing this pathway in the yeast cell itself has been suggested as an even better strategy as it

would leverage the yeast’s capability as an industrially robust microbe and also exploit the yeast’s

http://task39.org/files/2013/05/IEA-Task-39-Current-Status-and-Potential-of-Algal-biofuels.pdf

http://www.algaecluster.eu/

http://www.naabb.org/

http://www.syntheticgenomics.com/what/renewablefuels.html

http://www.syntheticgenomics.com/what/renewablefuels.html


July 2014 152

eukaryotic ability to more highly overexpress terpene (10-carbon isoprenoid-derived alkenes) synthases

which are usually of plant origin (e.g., Artemisia annua) and thus more difficult to express at high

activities in prokaryotes. The pioneering work of Ro et al. (2006) showed that overproduction of IPP and

DMAPP is possible in yeast and these researchers have engineered yeast cells to produce 115 mg/L of

artemisinic acid, a closer intermediate to the antimalarial drug artemisinin than amorphadiene. One of

the highest profile companies operating in this area is Amyris, which has leveraged its original research

which was focussed on developing artemisinin as an antimalarial drug. Amyris is now using further

derivatives of its modified microorganisms to produce farnesene, a diesel fuel precursor.

Figure 4-4: example isoprenoids

Other pathways such as the ketoacid pathway are also able to produce longer chain carbohydrates

(Jovanovic-Tews et al., 2010). They are not reviewed here, however, because at present they are

primarily applied to produce pharmaceutical and other specialty molecules, not biofuels.

4.3 Feasibility considerations

The promise of advanced biological pathways and processes for converting carbohydrate feedstocks to

biofuels or drop-in biofuels intermediates is that they can be “bolted on” to existing sugar or corn starch


July 2014 153

mills. The economic viability of such processes is primarily influenced by biological conversion

performance, efficiency and the amount of process inputs (e.g., nutrients, process energy and

equipment) required to produce and recover the target product molecules from aqueous production

broth. As sourcing sugar feedstocks as carbon sources for such processes is relatively costly, to achieve

good process economy the biofuel (or biofuel intermediate) production and recovery process steps must

be highly efficient and robust to achieve the type of performance demonstrated by the “gold standard”

of established ethanol fermentation processes. The efficiency of biological conversion processes is

typically characterized using the five measures described below:

Yield (mass of product per unit mass of feedstock): The proportion of carbon source feedstock

(e.g., sugars or biomass) converted into the desired product as opposed to less desirable products

such as cell mass, CO2, or non-target compounds. The minimum product yield required to achieve

cost competitive production is a strong function of feedstock cost.

Titer: The concentration (g/L) of product in the production medium at the completion of the

biological production process. The final product titer is strongly associated with process capital

and operating costs since lower titers typically mean more production vessels (or volume) are

required to produce a given quantity of product and greater amounts of process energy are

required to recover a given amount of product.

Productivity: The volumetric production rate (g/L/h) at which product is formed. Productivity

strongly influences capital expenditure since slower production rates mean a greater residence

time (e.g., more fermenter vessels) is required to achieve a given production target (e.g., t/d). For

example, doubling volumetric productivity results in a 50% reduction in capital costs per unit of

product.

Recovery: It is insufficient to simply generate a desired product, as it must also be recovered at

high yields and in usable form. The percent product recovery indicates how much of the

generated product is economically recoverable, especially for products that are accumulated

intracellularly. This is often well below 100%. For example, most fatty acids are generated

intracellularly and their recovery requires cells to be permeabilised or lysed, making high levels of

product recovery more techno-economically challenging.

Robustness: The biological production microorganism must be able to be used at an industrial

scale where there is less ability to precisely maintain targeted operating conditions (e.g., pH,

temperature, concentration of dissolved gases, etc.). Yeasts, especially S. cerevisiae strains that


July 2014 154

convert sugar feedstocks (hexoses in particular), are well recognized for their unrivalled industrial

robustness. This is why many companies pursuing commercialization of biological drop-in biofuels

technologies are proposing to heterologously express desirable metabolic pathways found in less

industrially-proven microbes such as Clostridia and E. coli bacteria into yeasts.

The volumetric productivity metric combines biological process production time and final product titer.

The volumetric productivity, expressed as mass of product per liter of reactor capacity per unit time

(g/L/h), represents an overall average productivity for the biological process and provides one good

indicator of its relative capital intensity. Higher volumetric productivity indicates that less reactor

volume is needed to produce the same amount of targeted product. A higher product titer generally

means a less costly product recovery system. The titer and volumetric productivity of selected biological

processes relevant for drop-in biofuel production are summarized in Table 4-1, with the corresponding

values for conventional ethanol fermentation processes.

Another important process metric is product yield. As noted earlier, ethanol processes are relatively

efficient in channeling a significant proportion of substrate carbon to product, routinely achieving 90–

95% of theoretical yield in converting 45 – 48% of feed sugar on a mass basis to ethanol. Lower

theoretical and mass yields can be expected for biological conversion processes that produce more

reduced molecules with higher Heff/C ratios, as proportionally more oxygen, which represents

approximately 46% of carbohydrate mass, must be removed. As a result, theoretical butanol mass yields

from carbohydrates via the ABE pathway are 41% compared to 51% for ethanol. In reality, primarily due

to product feedback inhibition, butanol mass yields from carbohydrates are in practice typically only 10-

15%, about 33% of theoretical (Ranjan & Moholkar, 2012). Similarly, theoretical mass yields of the highly

reduced molecule farnesene from glucose are estimated to be only 25 – 29% (Rude & Schirmer, 2009).

To date, achieved yields have been about 50-67% of this theoretical value, with Amyris reporting 17 wt%

farnesene yields from glucose (Pray, 2010) and the National Advanced Biofuels Consortium (NABC)

analysis of Amyris’ process reporting a 13% mass yield (Foust, 2011).

One notable feature of the Amyris process is that, according to the NABC, the carbon efficiency of the

process is in the order of 60% with the remaining 40% available for energy production (O’Connor, 2013).

While this results in low farnesene yields it can translate into a GHG credit that can be used as a positive

attribute for marketing the technology. Essentially this 40% carbon is in the form of non-farnesene

carbohydrates that end up in the anaerobic digester of the facility. Thus, for every unit of farnesene


July 2014 155

Table 4-1: Biological process performance metrics for selected biofuel production microorganisms

1more than 12% w/v in US conventional corn-to-ethanol mills

Product

(Fuel)

Production

Microbe

Fermentation

time

Titer

(g/L)

Volumetric

Productivity

(g/L/h)

Theoretical

Yield

(g/g glucose)*

Major start-up

companies

Major

technological

challnge

Reference

Ethanol Saccharomyces sp. 6-12h 47 – 55 (6-11% v/v)1

2.00-3.00 0.51 commercial Cost of sugar or starch feedstock

(Bai et al., 2008; Khosla, 2011)

Butanol Clostridium sp. 48-96h 8 -17 0.08-0.46 0.41 Butamax, Cobalt,

Gevo, Green biologics, etc.

End product inhibition - low titers

(Jang et al., 2012)

Butanol Clostridium sp. 72h 20 0.34 0.41 Idem Idem (Qureshi & Blaschek,

1999) Farnesene Amyris

Saccharomyces sp. 150h 104 0.70 0.25-0.29 Amyris Yield, productivity (Pray, 2010; Rude &

Schirmer, 2009) Ethanol Syngas fermentation

Clostridium ljungdahlii

n/a 11-48

n/a n/d LanzaTech, INEOS Bio, Coskata

Gas-liquid mass transfer/low water solubility of H2, CO

(Munasinghe & Khanal, 2010)

Fatty acids Dunaliella Salina n/a 2-6 0.3 g/m2/h 0.29 Solazyme Yield, recovery (Darzins et al., 2010)


July 2014 156

(diesel) energy, one unit of methane energy is produced. This methane is combusted to generate heat

and power (about 87 kWh of power per tonne of sugar according to O’Connor (2013)). As half the

recoverable energy (excluding carbon in CO2) from the sugar is stored in the diesel product and the

other half is utilized for combustion this results in a significant GHG savings credit to the overall process.

Amyris claims that the life cycle of their Biofene fuel emits 80% less GHG than petroleum diesel (Amyris,

2013b)

A significant degree of technological uncertainty and risk remains in biological processes for producing

drop-in biofuel-relevant molecules, especially when genetic engineering of the production

microorganisms is involved. The lipids and solvent-type product molecules used as precursors for drop-

in blendstocks are usually secondary metabolites which are not directly associated with cellular energy

metabolism but more with cell growth, maintenance and survival. They are typically produced in only

small amounts in excess of what is required for cell growth unless the microorganism is stressed.

Another challenge is that, when larger amounts of these types of products are produced, it is usually at

the expense of cell growth. For example, algae produce higher yields of lipids when they are nitrogen

starved and are thus limited in carrying out the protein synthesis needed for growth (Darzins et al.

2010). From a microbe survival standpoint, saturated compounds like lipids are accumulated in non-

growing cells as a means of storing energy to protect the organism against potential future energy

limitations (Darzins et al., 2010; Jovanovic Tews et al., 2010). There is also usually a lag in the time

between the initiation of the biological process and the onset of accumulation of secondary metabolites

like lipids. In contrast, ethanol production by yeast (or bacteria) occurs by a central metabolic pathway

linked to primary energy metabolism (Bai et al., 2008). As such, fermentative ethanol production is

directly associated with cell growth and there is no time lag in ethanol production after the process is

initiated.

The industrial robustness of microbial production strains and the ease of scale up of biological

production processes will also have a significant influence on how quickly advanced biological processes

for drop-in-relevant products can be successfully commercialized. The industrial challenges differ

signficantly for aerobic (isoprenoid and fatty acid) and anaerobic processes (n-butanol and isobutanol).

Although biobutanols can be produced in strictly anaerobic processes for which considerable industrial

experience has been developed over the past decades, butanol production processes so far continue to

exhibit relatively slow growth rates and low product titers, which makes them more susceptible to

contamination by competing microbes (as well as bacteriophages if bacterial production hosts are used).


July 2014 157

A recent techno-economic assessment comparing butanol and ethanol production found that, on an

energy basis (lower heating value, LHV), butanol had 50% of the yield and 25% of the productivity of

ethanol, with the lower yield translating into higher feedstock costs per unit of fuel energy produced

while the lower productivity translated into higher capital costs (Pfromm et al., 2010). No minimum fuel

selling price (MFSP) was reported in this study but other work using the US DOE standard large scale

process model basis (2000 tpd feedstock input, 10% return on investment) and a feedstock price of US

$65/t, projected the MFSP of butanol to be US $5-6/gal ($1.3-1.6/L) (Jovanovic-Tews et al., 2010), which

is slightly higher than the commodity chemical price for butanol (ca. USD $5.00/gal or $1.32/L, (Informa

Economics, 2013)) and significantly higher than the energy equivalent gasoline price (US $2.25/gal or

$0.60/L, based on USD $2.9/gal & 36 MJ/L gasoline and 28 MJ/L butanol).

The two aerobic processes discussed in this chapter, isoprenoid and fatty acid (lipid) synthesis, have not

yet been demonstrated to be industrially robust at large scale (> 500,000 L) thus their technology risk

and contingency allowances remain higher than currently projected for butanol (Amyris, 2013; LS9,

2013; Gevo, 2013). Another issue is that in order to produce the reducing power required to drive

biosynthesis a significant proportion of feed carbon source must be oxidized to CO2 a reaction which is

also exothermic. The exothermic nature of aerobic submerged biological production processes, as well

as the need to constantly supply molecular oxygen (O2) at sufficient levels to maintain aerobic

conditions, pose significant challenges to achieving economical scale up. Table 4-1 compares scale up

issues encountered in aerobic biological production and anaerobic fermentation processes.

Table 4-2:Scale up issues for anaerobic vs aerobic fermentation

Anaerobic Aerobic

Heat generation 235 KJ/mole 2840 KJ/mole Ability to dissipate heat externally Yes No Compressed air inputs Yes No Largest reactor size 3,700,000 L 1,430,000 L Average reactor size 2,800,000 L 380,000 L Risk of ceasing digestion due to high pressure Lower Higher

Source: (Hannon et al., 2007)

The challenging task of assessing the feasibility of biological (and thermochemical) drop-in biofuel

processes is being tackled by the US DOE-funded National Advanced Biofuels Consortium (NABC) (box 4-

1). Notably, the NABC has selected Amyris’s aerobic isoprenoid metabolism-based farnesene production

http://www.ls9.com/

http://www.ls9.com/

http://www.gevo.com/


July 2014 158

process as one of its two most promising technologies for scaling up commercial production of drop-in

biofuel precursors.

Another aspect of industrial scale up is the opportunity to leverage existing oil refinery infrastructure. In

the case of biological production processes, the only step that can readily be performed at a petroleum

refinery is the final hydrotreatment step, if needed, as well as perhaps final blending to create a fungible

blendstock. As long as the oxygen content of biologically produced intermediates can be brought to

roughly the same level as vegetable oils or fatty acids (~ 10 wt%), they can be treated in a similar way as

oleochemical feedstocks, as discussed in Chapter 2.

In addition to hydrotreatment, molecular hydrogen (H2) can also be used to boost reducing processes in

metabolic pathways and thereby improve yields. The more external hydrogen that can be used as a

source of “reducing” energy in biological production processes, the more carbon can be preserved in the

final fuel product. The higher the energy content and the lower the oxygen content of the product

molecules, the more energy is required by the production microorganism for its biosynthesis. For every

Box 4-1: NABC: The US DOE grand challenge for drop-in biofuel technologies

The National Advanced Biofuels Consortium (NABC) is a

partnership of 17 industry, national laboratory, and university members to develop technologies to

convert lignocellulosic biomass feedstocks to biofuels that are compatible with the existing US

transportation infrastructure. With $35 million of American Recovery and Reinvestment Act funding

from the U.S. Department of Energy and $14.5 million of partner funds, the National Advanced

Biofuels Consortium program started in 2011 and is due for completion by the end of 2013. In stage I

(early 2011), NABC examined six process routes spanning biochemical, thermochemical and hybrid

approaches to assess their potential to meet project objectives. In Stage II, the Amyris biochemical

and Virent hybrid (catalytic sugar processing technology, discussed in Chapter 5) process routes

were selected for further research and development. Two thermochemical routes – hydropyrolysis

and hydrothermal liquefaction – were refocused to address key process issues. The consortium

continues to develop and assess the feasibility of these downselected processes (NABC, 2013).


July 2014 159

step of decarboxylation or reduction of metabolic intermediates, some fermentable carbon is consumed

and oxidised to CO2 in order to provide the necessary reducing power (in the form on NADPH or HCO3)

for the reductive biosynthesis reactions to occur. In other words, making energy-dense and low oxygen

content biohydrocarbons via biological processes consumes extra carbon and thus reduces yields. As is

emphasized in the next section, Chapter 5 on hybrid conversion platforms, the use of external hydrogen

inputs can be employed in syngas fermentation, particularly when the syngas source is hydrogen

deficient. For example, LanzaTech’s biological approach is able to use CO2, the most oxidized form of

carbon, but only if significant molecular hydrogen is also provided.

4.4 Commercialization efforts

A number of companies are trying to commercialize the biochemical route to drop-in biofuels and some

examples are summarized below.

4.4.1 Butanol

Butamax, a joint venture between Dupont and BP, is focused on isobutanol

production by engineered microbes. It is operating in a similar technology

space as Gevo (see below) and the two companies are engaged in on-going

litigation over intellectual property. One of Butamax’s patents is on

isobutanol production in yeast (Feldman et al., 2008) and another one on

increasing dihydroxyacid dihydratase (DHAD) activity (Dundon et al., 2011) to improve production of

isobutanol and other metabolites in yeast. Butamax has a demonstration unit in Hull, UK. They

ultimately plan to utilize the 190 MLPY (50 MGPY) High Water corn ethanol facility in Minnesota, USA for

commercial production (Lane, 2011a). In November 2012, Butamax added a technology laboratory in

Paulina, Brazil in order to further test its biosynthetic isobutanol production technology using sugarcane

derived sucrose feedstock (Biofuels-Journal, 2010).

Gevo, based in Colorado, USA, is trying to commercialize their

patented Gevo Integrated Fermentation Technology (GIFT)


July 2014 160

approach to isobutanol production. Unlike n-butanol, isobutanol is branched and more suitable for fuel

blends (as discussed in section 4.2.1). The company claims it can achieve higher yields and recoveries of

isobutanol compared to n-butanol. The GIFT process consists of genetically engineered yeast and

continuous in situ recovery tailored to isobutanol extraction from fermentation broth. Although details

are not publically disclosed, based on publically available presentations and patents (Evanko et al.,

2012), the preferred recovery route is likely to be gas stripping because isobutanol has a boiling point of

108 °C compared to n-butanol’s boiling point of 118oC. This attribute in itself possibly makes the

recovery less energy intensive. The exact productivity of the GIFT platform is not publically known

although in consultancy reports Gevo has claimed 5 times higher productivity than reported to date for

n-butanol (Nexant, 2012). Assuming that n-butanol’s practical volumetric productivity is 0.3 g/L/h, the

GIFT platform should achieve 1.5 g/L/h. This rate is half the current productivity of ethanol as

summarized in Table 4-1. Gevo has a strategic collaboration with ICM, the engineering company that has

designed 60% of the US’s existing corn to ethanol mills. The first 3.8 MLPY (1 MGPY) demonstration

facility was “bolted on” to a corn to ethanol facility in St Joseph, Missouri. In September 2010, Gevo

acquired its first production facility in Luverne, Minnesota. In early August 2013, Gevo announced that

its second, one million liter fermenter and GIFT system was brought online at its Luverne, Minnesota

facility, and they also announced they would be opening a facility for catalytically upgrading isobutanol

to renewable paraxylene and jet fuel in Texas in September 2013. Other strategic partnerships have

been formed, including with Mustang engineering for jet fuel tests and certifications and with Cargill to

develop a yeast biocatalyst for cellulosic feedstock derived isobutanol.

While Gevo and Butamax are targeting isobutanol, many other biobutanol start-ups are focused on

“conventional” n-butanol produced via the traditional ABE pathway. Cobalt, Green Biologics, GranBio,

Microvi, Optinol and Rhodia are all examples of companies working to commercialize n-butanol

production. Several of these companies intend to retrofit existing sugar or corn mills for butanol

fermentation and recovery. Most companies claim they will be able to achieve cost parity with ethanol

on an energy basis.


July 2014 161

4.4.2 Fatty acids, long chain alcohols and alkenes

Amyris is a University of California, Berkeley spin off company

founded in the San Francisco Bay area in California, USA in 2003. In

2005, the company received a grant from the Bill and Melinda Gates

Foundation to optimize a yeast-based isoprenoid biosynthetic platform for producing the antimalarial

drug precursor artemisinic acid (Amyris, 2013). Using this same synthetic biology platform Amyris later

collaborated with the French oil company Total to commercialize production of farnesene as a diesel

and jet drop-in biofuel component. Farnesene, marketed by Amyris under the trade name BiofeneTM, is

also used as a precursor for cosmetic ingredients such as squalene which is normally extracted from the

fat of large sea mammals. The biofuel application of Biofene derivatives such as the alkane farnesane

were tested in blends of diesel for automobiles and in blends of jet fuel for aircraft at the 2013 Paris Air

Show (Amyris, 2013a). The blending ratio of Amyris biofuel in the petroleum jet fuel was not specified.

LS9 is another San Francisco Bay area company that has developed a

proprietary synthetic biology platform based on E. coli, with long chain fatty

alcohols currently its main product focus (LS9, 2013). By leveraging E. coli’s

natural ability to metabolize a range of hexose and pentose sugars, the company claims they are well

positioned to utilize lignocellulosic sugars and provide a “capital-light” production platform (LS9, 2013).

They have partnered with Virdia, a company targeting biomass sugars production, on a $9 million US

DOE grant to build a demonstration plant in Florida, USA to convert woody biomass to fuels. However,

the facility will initially use glucose (starch derived sugar) as the feedstock. In 2010, LS9 purchased a

demonstration facility in Florida, and has since successfully demonstrated its production technology in a

135,000 L fermenter (LS9, 2013). In the short term, LS9 is primarily focusing its technology on higher

value chemicals rather than biofuels.

Solazyme is another San Francisco Bay Area company which has

developed a proprietary biotechnology platform based on heterotrophic

algae. Using sugar feedstock Solazyme’s algae are able to make a range of

biological oils (TAGs), with various products sold in cosmetics stores. AlgenistTM, for example, is a line of

http://www.amyris.com/News/311/Amyris-and-Total-Announce-Successful-Demonstration-Flight-With-Renewable-Jet-Fuel-During-Paris-Air-S

http://www.amyris.com/News/311/Amyris-and-Total-Announce-Successful-Demonstration-Flight-With-Renewable-Jet-Fuel-During-Paris-Air-S


July 2014 162

anti-aging skincare products formulated with alguronic acid sold in about 1000 stores across the US,

Canada and the UK (Solazyme, 2013). The company has stated that its strategy is to first develop the

skin care, cosmetics and other value-added chemicals markets and then to scale up production to supply

transport fuel markets. In 2010, the company sold 80,000 L of hydrotreated algal oils to the US Navy for

use in marine fuel blends. These oils cost several times more than their equivalent petroleum based

counterparts, as profiled in several media outlets (Alexander, 2012; Roig, 2012). However, the company

claims that their oils can be priced in the same range as heating oils which at the time of writing is about

1000 USD/t (IndexMundi, 2013; Solazyme, 2013)

4.5 The value added chemicals opportunity

Current trends in biochemically based drop-in biofuel technologies suggest that the production of higher

value chemicals is a more attractive value proposition than biofuels, at least in the near term. Although

biochemically based drop-in biofuels have the potential to “piggy-back” on existing sugar or starch-to-

ethanol facilities, the lower volumetric productivities and yields of these still evolving industrial

processes is proving challenging when trying to achieve cost competitive biofuel production. In fact,

some biochemical products can be used either as biofuels or as chemical feedstocks or solvents. Thus,

the biochemical based platforms have an advantage in that they can be used to also produce higher

value chemicals at high selectivities (See Box 4-2 describing NatureWorks approach). As the use of these

products as chemical feedstocks tends to be more profitable than as fuels, it is unlikely that most

biochemical drop-in production will be used as fuels before their higher value markets as chemicals are

saturated. Normal butanol, for example, commands a price of ca. USD $5.00/gal or $1.32/L, (Informa

Economics, 2013) as a commodity chemical versus US $2.25/gal or $0.60/L (based on USD $2.9/gal & 36

MJ/L gasoline and 28 MJ/L butanol) as a gasoline energy equivalent. As biochemical product chain

length gets longer even more valuable applications are possible. For example, as Amyris has

demonstrated, longer chain lipids are likely to command higher prices in the cosmetics industry.

From an energy and hydrogen balance standpoint, biochemical platforms are more efficient at

producing oxidized products than highly reduced molecules. Highly oxidized products such as lactic,

acetic and adipic acids command significantly higher prices than ethanol or biodiesel and they are

challenging to make by non-biochemical platforms. Such oxidized molecules can also be produced by

oxidation of petroleum cuts using various processes (e.g., the oxosynthesis process for butanol). While


July 2014 163

technically feasible they are costly and entail the risk of over-oxidizing and thus “burning” up a portion

of the petroleum feed and incurring associated yield losses (Gary et al., 2007; Vennestrøm et al., 2011).

Since sugar molecules inherently contain high amounts of oxygen, it is more efficient to convert them to

oxidized intermediates and aim for higher value chemicals markets. The Heff/C ratio staircase shown in

Figure 4-5 demonstrates the number of steps that have to be taken to bring sugar feedstocks to the

functional equivalence of petroleum products. It is apparent that fewer steps must be ‘climbed’ (or none

in the case of sugar conversion to acetic or lactic acids) along the Heff/C ‘staircase’ for sugar to be

converted to intermediates that, in the short term at least, can be sold as chemical precursors rather

than as fuels. Synthetic biology is advancing rapidly and new tools provide ever greater flexibility as to

the range of products and their tuning for desired customized chemical properties that can be achieved

through biochemical platform routes. The ability to tailor complex molecules and achieve extremely high

production selectivity is generally absent in thermochemical technologies.

Figure 4-5: The H/Ceff ratio in relation to the biochemical platform and its intermediates

An example of a biofuel process that produces intermediates that are of potentially higher value as

chemicals than fuels is the technology proposed by Zeachem. The company uses the acetogenic

bacterium Clostridium thermoaceticum to convert sugars to acetic acid (a more oxygenated compound

Sugar 0

0.2 Wood

Diesel 2.0

1.8 Lipids Biochemical feedstock

‘Drop-in’ biofuels

1.0

0.6

0.4

0.8

1.2

1.4

1.6

Fatty acids

Farnesene


July 2014 164

than ethanol, see staircase in Figure 4-5) and the acetic acid is then recovered (in the form of acetate

salts) and hydrogenated to ethanol (Zeachem, 2013). Hydrogen for hydrogenation is imported, although

in theory it can also be generated from biomass lignin residues. The biggest advantage of this acetogenic

pathway is that all carbon is funneled to acetic acid formation and there is no carbon “waste” funneled

to the formation of CO2. Although this platform can be used to produce biofuels, the acetic acid it

produces is too oxidized to be a drop-in biofuel intermediate, and the hydrogen required to convert it to

ethanol may outweigh the economic and environmental benefits of avoiding CO2 production. At the

time of writing, the value of acetic acid (USD $3000/t) is higher than the net revenue that would be

achieved by hydrogenating it to ethanol (USD $850/t).

Another notable example of sugar feedstocks being used to produce higher value oxygenated chemical

intermediates, such as polylactic acid (PLA) for polymer applications is NatureWorks, which was

originally formed as a joint venture between Cargill and Dow. Natureworks primary manufacturing

facility in Blair, Nebraska has the capacity to produce 140,000 t/year of PLA from corn starch derived

glucose. PLA polymers have many uses in textiles and plastics markets, including as renewable packaging

materials (see Box 4-2). Sugar conversion to lactic acid by fermentation exhibits a theoretical mass yield

from glucose of 1.00 and process production yields of up to 0.9 g/g are reported (Hofvendahl & Hahn-

Hägerdal, 2000).

The counterargument for using biological processes to convert renewable feedstocks to higher value

chemicals instead of fuels is that the markets for chemical products are quite limited compared to those

for fuels. Although this is true, bio-based chemical markets are growing rapidly and expected to

continue to expand for the foreseeable future, largely as a result of a sustained increase in petroleum

prices and growing awareness of the benefits of being able to market renewably sourced products.

Several of the advanced biological conversion companies reviewed earlier (e.g., Amyris, Gevo, LS9,

Solazyme, etc.) are currently generating a significant part of their revenue from sales of biobased

products rather than drop-in biofuels (Amyris, 2013a; Bomgardner, 2012; Gevo, 2013; LS9, 2013;

Solazyme, 2013). Other companies have clearly stated they are primarily targeting value added

oxygenated products such as dicarboxylic acids (e.g., Myriant’s succinic acid) and dialcohols (e.g.,

Genomatica’s butanediol or DuPont’s 1,3-Propanediol). A 2012 issue of E&EN magazine provides a world

map showing the breadth of biochemically based facilities being built for the production of chemical

feedstocks (Bomgardner, 2012). Most of these facilities are located in biofuel producing countries such

as the US and Brazil.


July 2014 165

A number of business intelligence organizations have also published estimates of the future size and

growth rate of biobased chemicals markets. For example, Nexant projects that by 2015 global biobased

chemical markets will grow to 5 million metric tonnes per year (Bomgardner, 2012) while Lux Research

projects that biobased materials markets will grow 17.7% per year to reach 8.1 million tonnes by 2015

(Lux Research, 2010). The UK’s National Centre for Biorenewable Energy, Fuels and Materials (NNFCC)

estimates that by 2020 the global market for renewable chemicals and polymers will reach 50 million

metric tonnes and that oxygenated fermentation products such as alcohols and carboxylic acids will

account for 17% of this market (i.e., 8.5 million metric tonnes) (Higson, 2011).

The Forest Products Association of Canada recently published its “Biopathways” strategic analysis, which

assessed various scenarios for adding value to Canada’s forest products industry (FPAC, 2011). A key

Box 4-2: NatureWorks, LLC: Biodegradable sugar-derived polymers at commercial scale

Nature Works LLC is a joint venture between Cargill and Dow, currently producing an annual total of

140,000 tonnes of polylactic acid (PLA). This polymer is produced at a dedicated facility in Blair,

Nebraska, USA using corn dextrose as the feedstock. The dextrose is fermented by Lactic Acid

Bacteria (LAB) and the resulting lactic acid is then polymerized to form various types of PLA polymers

for use in textiles and plastics formulations. The facility converts $400-500/t dextrose to $3000-

4000/t PLA (approximate global prices at the time of writing, www.alibaba.com). The company was

founded in 1997 as Cargill Dow LLC with original plans to begin operations in 2001. The company was

renamed NatureWorks LLC in 2003 and full operations at the Nebraska facility began in 2009. The

PLA product is marketed as IngeoTM and its production is reported to result in significantly lower

GHG emissions than petroleum derived equivalents (Binder & Woods, 2009). The polymer is

biodegradable and degrades without generating significant amounts of methane (Kolstad et al.,

2012). The 2012 London Olympics used 8,000 t of IngeoTM for food and other packaging, which after

use was composted to produce soil amendments.

http://www.alibaba.com/


July 2014 166

conclusion was that value added renewable chemicals should be prioritized to help modernize the

Industry, diversify its markets and improve its profits. As an example, the bio-plastics and plastic resins

market is expected to grow by 23.7% a year to reach USD $3.6 billion market by 2015 (FPAC, 2011).

Although these markets are still quite small compared to transport fuels markets of 2 billion tonnes and

USD $3 trillion, they are sizeable compared to the current capacity of the biomass processing sector and

of similar scale to the current biofuel production volume of 60 million tonnes (IEA, 2012b).

In summary, the biochemical platform, in contrast to the oleochemical and thermochemical platforms,

can produce pure molecular streams of functionalized molecules which currently command higher

market prices in chemical markets than as intermediates for drop-in biofuels. The primary role of the

metabolic pathways involved in drop-in biofuel production is to biologically deoxygenate the sugar

feedstock (low Heff/C) to produce a more highly reduced or saturated products such as isoprenoids and

fatty acids (high Heff/C). These biological processes are in general comparatively energy- and carbon-

intensive and achieve productivities about one order of magnitude lower than conventional sugar-to-

ethanol fermentation processes. Despite this situation, interesting possibilities exist to increase rates

and yields of biochemical based processes by inserting externally supplied hydrogen or otherwise supply

the needed reducing power. The addition of external hydrogen could be used to minimize the need for

microbes to generate their own reducing power by oxidizing a portion of the sugar, starch or cellulose

carbon feedstock. However, hydrogen inputs are costly and unlikely to be used in biochemical platforms

to produce commodity drop-in intermediates.

With fewer processing steps and less hydrogen consumption and higher feedstock conversion yields,

biochemical platform technologies can be used to produce oxygenated and low Heff/C products that will

likely generate higher revenues as higher value chemicals than as biofuels. Market saturation issues for

renewable chemical products should not be too much of a concern in the near term due to projected

strong growth in bio-based material markets.


July 2014 167

CHAPTER 5: HYBRID PLATFORMS

Although the majority of drop-in biofuel technologies are captured under one of the previously

described oleochemical, thermochemical or biochemical platforms, some technologies combine

approaches from two or more platforms and are thus referred to as “hybrid” routes to drop-in biofuels.

The main examples include:

Biological conversion (fermentation) of biomass gasification-derived syngas to alcohols.

Companies that have investigated or are developing this approach include Coskata, LanzaTech

and INEOS Bio.

The thermochemical catalytic conversion of biochemical platform-produced sugars and perhaps

other solubilized carbon species to hydrocarbon biofuels. The company Virent Energy Systems

(“Virent”) is the primary current developer trying to commercialize this approach.

The thermochemical catalytic conversion of biochemical platform produced alcohols (e.g.,

ethanol or butanol) to hydrocarbon biofuels. Companies pursuing this approach include BIOGY,

Cobalt and Gevo.

Each of these hybrid technologies offer advantages as well as present new challenges compared to

thermochemical or biochemical platform routes discussed in chapters 3 and 4, respectively. For

example, traditional sugar fermentation is able to ferment only the saccharide portion of biomass

whereas syngas can be produced from the entire biomass carbon (i.e., including lignins, extractives and

other non-sugar carbon-containing compounds) such that there is potential to convert more of the

feedstock to fuel product(s) via syngas fermentation. Similarly, catalytic conversion of sugars and other

organic carbon compounds offers the potential to convert sugars and perhaps more than just sugars to a

broader range of higher energy products than available via traditional biological conversion

technologies. The third approach, alcohol-to-drop-in, involves the catalytic conversion of higher Heff/C

alcohols rather than lower Heff/C sugars (for which Heff/C = 0) but is otherwise conceptually similar to

catalytic upgrading of sugars. These three hybrid routes to drop-in biofuels or fuel blendstocks are

discussed further in the following sections.


July 2014 168

5.1 Syngas fermentation

As described in Chapter 3, the conversion of syngas to liquid biofuels can be achieved thermo-

catalytically. However, some fermentative microorganisms can also utilise syngas as their carbon and

energy source while producing potential drop-in biofuel intermediates as a major product. Some

autotrophic microorganisms can use single carbon compounds such as CO and CO2 as sources of carbon

and energy-rich compounds such as CO and H2 as sources of energy. In the same way thermochemical

syngas upgrading involves the use of catalysts, microorganisms that convert syngas to biofuels such as

ethanol require metals such as cobalt and nickel (in relatively trace amounts) to catalyze essential

metabolic enzyme activities (Hayes, 2013). A range of microorganisms can ferment syngas to

intermediates such as ethanol, butanol and acetic acid, including acetogenic bacteria such as Clostridium

ljungdahlii and Clostridium carboxidivorans (Munasinghe & Khanal, 2010). These microorganisms use

the Wood-Ljungdahl pathway to reduce CO to Acetyl-CoA, from which ethanol and butanol are

produced via the pathways described in Chapter 4. The reducing power required for the first reaction

comes from the H2 in the syngas or from the oxidation of CO to CO2 via a carbon monoxide

dehydrogenase enzyme (Daniell et al., 2012; Weber et al., 2010). Here again, as seen in thermochemical

deoxygenation routes discussed in Chapter 3, there is a trade-off between carbon efficiency and

hydrogen consumption. From an evolutionary perspective these metabolic pathways are found in some

of the most primitive microorganisms on earth. Approximately 4 billion years ago, well before oxygen

became abundant in the atmosphere, these types of microorganisms survived by metabolising CO and

H2 gases (Daniell et al., 2012).

Syngas fermentation is claimed to have several advantages compared to its “parent” processes, i.e.,

sugar fermentation and thermocatalytic syngas conversion (Daniell et al., 2012; Munasinghe & Khanal,

2010). For example, compared to sugar fermentation, this approach is able to utilise lignin in addition to

carbohydrate fractions of biomass. In comparison to thermocatalytic syngas conversion it has also been

claimed to be economical at a smaller scale, because of lower capital costs, while proving to be less

sensitive to impurities (Daniell et al., 2012). However, the volumetric productivity of this hybrid

approach is still quite low, as summarized in Table 4-1, which shows current volumetric productivities

are around 0.3 g/L/h, similar to the biochemical production of butanol and farnesene. The primary

limiting step for syngas fermentation is the low solubility of CO and H2 gases in aqueous solutions as

these gases must first be absorbed into the aqueous fermentation broth to be microbially assimilated


July 2014 169

(Bradwell et al., 1999). As the gas-liquid mass transfer and solubility of both molecular hydrogen and

carbon monoxide in water are quite low compared to more conventional sugar substrates (where gas-

liquid mass transfer isn’t needed), syngas fermentation rates are typically constrained by physical mass

transfer limitations rather than being limited by metabolic capacities. From an energy balance

perspective, syngas also has to be cooled down from a production temperature of ca. 700 °C or higher to

ca. 50 °C or less before it can be fermented. This significant cooling requirement is of concern because,

as discussed in Chapter 3, large temperature fluctuations in gasification systems result in significant

engineering challenges and generally have an adverse effect on process energy efficiency.

The gas-liquid mass transfer challenge of syngas fermentation platforms has been a long-standing focus

of engineering research and various reactor designs have been assessed to try to improve the contact

surface area between gases and liquids and achieve increased dissolution and diffusion rates. Designs

include continuous stirred tank, microbubble dispersion stirred tank, bubble or gas lift columns, hollow

fiber and other membrane-based systems, trickle bed and monolithic biofilm reactors. A detailed

description of possible reactor designs and their operation and relative efficiencies is provided by

Munasinghe & Khanal (2010).

Several companies and research institutions are trying to develop and commercialize drop-in biofuels

based on variations of the syngas fermentation technology platform. Some of the leading companies

and their recent activities to commercialize this hybrid route are briefly described below:

INEOS Bio is a fully owned subsidiary of the oil and chemicals company INEOS. Their

technology builds on the pioneering development of syngas fermentation technology

carried out by James Gaddy and colleagues at the University of Arkansas and then

Bioengineering Resources beginning in the 1980s (Bryan, 2011; INEOS, 2013). INEOS Bio’s technology is

based on a patented process using proprietary microorganisms to convert biomass derived syngas to

ethanol and other bio chemicals. INEOS Bio recently completed a 30 MLPY (8 MGPY) syngas to ethanol

facility in Vero Beach, Florida, USA. This facility, named the Indian River Bioenergy Center, started

producing and commercially selling ethanol in August 2013. The plant, which cost $130 million to

construct (backed by $75 million in US DOE loan guarantees), uses as feedstock municipal solid waste

(MSW) from the Vero Beach municipality. Besides producing ethanol, the facility is designed to


July 2014 170

coproduce 6 MW (gross) of power. Facilities capable of using MSW feedstock can have a significant

competitive economic advantage because they essentially operate as a waste management facility and

receive a tipping fee for taking in feedstock and thus are producing products from a negative cost

feedstock. The 6 MW power generation component is important to the economics of the Indian River

facility as their power can be sold at a relative premium compared to many other jurisdictions in the

USA (INEOS, 2013).

Founded in New Zealand in 2005, and headquartered in

Chicago, Illinois, LanzaTech has developed a gas-substrate

fermentation process to produce ethanol and other and

chemicals from carbonaceous CO-rich gases. LanzaTech operates research laboratories in the USA, China

and New Zealand, and currently operates two sub-commercial scale facilities (Lanzatech, 2013). Their

pilot facility, located at the BlueScope Steel mill in New Zealand and linked directly to the mill’s off-gas

exhaust, has been operating since 2008 and has an ethanol production capacity of 15,000 gal/year

(56,800 L/year). In November 2012, LanzaTech completed the first phase of a multi-phase project with

Baosteel, China’s largest steel producer. This demonstration facility has an ethanol production capacity

of 100,000 gal /year (380,000 L/year) and will convert CO-rich waste gas from Baosteel’s production

facility using LanzaTech’s gas fermentation technology. The demonstration facility is intended to be a

precursor of a 190 MLPY (50 MGPY) commercial facility to be built in 2014. In 2012, construction of a

second demonstration facility was initiated near Beijing in partnership with Capital Steel, China’s fourth

largest steel producer. LanzaTech also plans to operate a commercial facility in Soperton, Georgia, USA

to be named the Freedom Pines Biorefinery based on its January 2012 acquisition of the former Range

Fuels biomass gasification facility (described in Chapter 3). The company has a number of high profile

partners including Siemens, Petronas and Baosteel on the production side and Virgin Atlantic and Boeing

on the product off-take side. For its aviation fuel projects, LanzaTech has partnered with Swedish

Biofuels AB which specializes in alcohol-to-jet catalytic conversion technology. All of the plants, other

than the Soperton facility, will use CO-rich waste gases from either the steel or oil refining industries as

their carbon source feedstock (LanzaTech, 2013).

LanzaTech recently announced a new process pathway to convert CO2 to ethanol using its proprietary

microorganisms (LanzaTech, 2013). Although this process has received a considerable amount of press

attention as a potential carbon capture solution, significant amounts of hydrogen will be required to

convert this fully oxidized form of carbon to a chemically reduced alcohol product.


July 2014 171

Syngas fermentation companies have shown ethanol can be commercially produced from CO and H2-rich

gases (syngas) or CO-rich gases via fermentation. Although ethanol is itself not a drop-in biofuel, at the

right price point it can be an attractive intermediate for drop-in fuel production. Moreover, the prospect

holds that similar fermentative pathways as used in gas-based ethanol production can be used to

produce even more reduced products, for example alcohols such as butanol, which are better suited

than ethanol as drop-in biofuel intermediates.

5.2 The Zeachem process

Zeachem’s process is based on using an acetogenic bacterium

isolated from the digestive system of termites to convert sugars

to acetic acid at near 100% carbon efficiency. Unlike sugar

fermentation to ethanol, in which one molecule of CO2 is lost for every molecule of ethanol formed, no

CO2 is produced in Zeachem’s sugar to acetic acid fermentation process. The acetic acid produced is

subsequently reduced to ethanol through the addition of hydrogen. Zeachem claims higher ethanol

production yields are possible using this “indirect ethanol” approach than in conventional “direct” sugar

to ethanol fermentations (Zeachem, 2013). However, the need to produce hydrogen, nominally by

gasification of biomass or lignin residues, introduces additional capital costs and operational

complexities. The Zeachem process represents a hybrid platform that uses fermentation to produce an

acetic acid intermediate and gasification to generate hydrogen to chemically reduce acetic acid to

ethanol or other chemicals or biofuel intermediates. The company currently operates a 3.8 MLPY (1

MGPY) ethanol demonstration facility in Boardman, Oregon, USA and sees potential to expand on its

technology platform to produce more highly reduced C3 and C4 products that could be used as drop-in

biofuel intermediates (Zeachem, 2013).

5.3 Aqueous phase reforming and the Virent process

The technology for catalytically converting sugars or polyol feedstocks to monooxygenates, also known

as aqueous phase reforming (or APR), is being developed by the company Virent Energy Systems

(Virent), which intends to use it to produce drop-in fuels and value-added renewable hydrocarbon

chemicals (Blommel & Cortright, 2008). This technology is classified as a hybrid platform because it


July 2014 172

converts sugars, nominally a biochemical platform feedstock, to drop-in biofuels using a thermochemical

catalytic process. Variations of APR also have been used to produce hydrogen from the aqueous

residues of pyrolysis oil separation and upgrading processes (Bridgwater, 2012). The APR technology

builds on the pioneering work of researchers at the University of Wisconsin who showed that highly

reduced carbon molecules could be produced when sugars (or other polyols) in aqueous solution were

mixed with molecular hydrogen and passed over ZSM-5 catalysts (Cortright et al., 2002; Huber et al.,

2004; Huber & Dumesic, 2006). This initial work ultimately lead to the creation of Virent and the

development of its patented BioFormingTM process, which comprises a combination of the original APR

process combined with upgrading technologies used in oil refineries, as further refined by Virent. In

overview, this process employs heterogeneous catalysts operating at moderate temperatures (175 to

300 °C) and pressures (10 to 90 bar) and reduces the oxygen content of the (plant biomass derived)

sugar or polyol feedstock through the three main steps described below (Blommel & Cortright, 2008;

Hayes, 2013; Huber & Dumesic, 2006):

a) Acid-catalyzed dehydration of sugars;

b) Aldol condensation of dehydrated carbohydrates over solid base catalysts to form large organic

molecules;

c) Dehydration/dehydrogenation of the large organic molecules to alkanes using bifunctional

catalysts (this step represents a variation of conventional catalytic processing as found in

petroleum refineries)

A schematic representation of the process is shown in Figure 5-1.

The Virent process yields a product exhibiting similar properties to petroleum reformate, although the

exact mixture composition depends on the nature of the catalysts and reaction conditions used. Acidic

catalysts such aluminosilicates (e.g., ZSM-5) favour the production of shorter molecules of more

phenolic nature (more suitable for gasoline blends) while base catalysts favour chain elongation through

aldol condensation and thus form longer chain liquid products (more suitable for diesel blends)

(Blommel & Cortright, 2008). Noble metal catalysts such as Pt and Rh favour reforming of carbohydrates

to hydrogen and CO2 (Bauldreay et al., 2010). The reformate can then be upgraded to a drop-in biofuel

using conventional refinery processes. The process consumes 12 wt% (of dry feedstock) hydrogen, some

of which is produced via the APR process itself, as shown in Figure 5-1. Virent claims that >90% of the

lower heating value (LHV) in the sugar and the hydrogen inputs is recovered in deoxygenated


July 2014 173

hydrocarbon products. Note, however, that there is so far limited public information on technical or

economic details of Virent’s process.

a)

b)

Figure 5-1: The Virent process. (a) Entire process. (b) simplified representation emphasizing key steps. Source: (Bauldreay et al., 2010; Blommel & Cortright, 2008)

Compared to fermentation, the APR approach is attractive because it has the potential to be used to

produce highly reduced (high Heff/C) hydrocarbons from a diversity of water soluble organic carbon


July 2014 174

compounds including both C5 and C6 carbohydrates and at much faster reaction rates than are possible

using biochemical routes (e.g., residence times of minutes vs hours or days). However, the APR reactions

are less selective than fermentation processes and produce a complex mixture of organic molecules.

Although the APR process uses non-exotic process components and catalysts, it faces similar catalyst

coking and deactivation challenges to those previously discussed in chapter 2 for pyrolysis oil upgrading.

This is because during catalytic upgrading of water soluble saccharides and bio-oil fractions, catalyst

coking and deactivation have been shown to increase with decreasing feedstock Heff/C ratio; feedstocks

with an Heff/C effective ratio less than 0.15 have been shown to cause large amounts of coke formation

(more than 12 wt%) from homogeneous decomposition reactions Zhang et al. (2011). Given that sugars

have an Heff/C effective ratio of 0, demonstrating the ability to avoid coking problems through effective

catalyst regeneration schemes will be key to commercializing aqueous phase reforming technology.

Headquartered in Madison, Wisconsin, USA, Virent was founded in

2002 with a goal of producing fuels and chemicals from renewable

sugars and other water soluble organic carbon compounds. Virent has

a 10,000 GPY pilot facility in Madison, built in collaboration with Shell,

one of Virent`s major investors. Other investors include Cargill and Honda. Other funding sources

include the US DOE, USDA and the US Dept. of Commerce (Virent, 2013). The Virent process was

selected (alongside the Amyris process) by the US DOE NABC program (see Chapter 4) as a promising

advanced biofuel platform to further develop for scale up to commercial production. In collaboration

with Coca-Cola, Virent is also trying to commercialize a proprietary process to convert sugars to

paraxylene (branded BioFormPX) (Virent, 2013).

5.4 The alcohol to jet process (ATJ)

As the name suggests, the alcohol-to-jet (ATJ) process converts ethanol or other alcohols such as

butanol to hydrocarbon mixtures which can be used as drop-in biofuels intermediates or blendstocks. In

the US, the ATJ process is receiving increasing attention both as a potential solution to the ethanol

blendwall as well as to possible production of ethanol beyond RFS mandated volumes.

The ATJ process is relatively simple, resembling the methanol to gasoline process discussed previously in

the syngas upgrading section of Chapter 3. The main processes, which have all been performed in the


July 2014 175

chemical and fuel industries for decades, involve dehydration, oligomerization and hydrogenation.

Although the technology risk is low and the ethanol feedstock costs can be estimated, it will likely be

economically challenging as it is an energy intensive process involving highly exothermic reactions.

Experience based on established methanol-to-gasoline processes (Shell’s MTG, see Chapter 3) have

shown that alcohol to hydrocarbon reactions are difficult to scale up due to the requirement for

multiple reactors and extensive heat exchanger installations (Allen et al., 1986). However, the

exothermic heat of reaction for ethanol-to-gasoline conversion is smaller than that of methanol to

gasoline. At an average value of 450 Btu per pound of hydrocarbons produced from pure ethanol, the

amount of heat to be dissipated will be about 3.7 times smaller than that from methanol (Allen et al.,

1986). However, another challenge, as discussed for the oleochemical platform, is the final drop-in

product might command a lower price than the feedstock alcohol.

The use of externally supplied hydrogen can be used to help increase the product yield although this

increases operating costs. Some proposed processes do not use hydrogen but generally result in lower

yields. The Oak Ridge National Laboratory (ORNL) in the US has recently developed a process that

operates without hydrogen inputs and at moderate reaction conditions (atmospheric pressure and 310

°C) to convert hydrous ethanol to hydrocarbons at yields reported to be over 50 wt% (Narula et al.,

2012). Preliminary techno-economic analysis of this process by NREL indicates that a minimum fuel

selling price (MFSP) of $3-4 per gasoline gallon ($0.8-1/L) can be achieved (Narula, Davison, & Keller,

2012; Davison, pers. comm. 2013). This work is still at a relatively early stage and other than this

preliminary techno-economic analysis little public information is available on the performance and

economic attributes of ATJ biofuel processes. However, NREL and MIT are in the process of conducting

separate techno-economic analyses of ATJ technologies that are expected to be published soon

(Pearlson, pers. comm., 2013).

As discussed in Chapter 2, ASTM approval for ATJ renewable jet fuel is expected sometime in early 2014

and this should further facilitate the commercialization of ATJ technologies. In anticipation of this

approval, large airline companies and the US Dept. of Defense have offered support to companies which

produce jet fuels from alcohols. Virgin Atlantic is supporting Lanzatech and Swedish Biofuels AB in their

goal of converting Lanzatech’s bioethanol to jet fuels (Lanzatech, 2013). Qatar Airways has invested in

the California-based alcohol-to-jet company BYOGY (Byogy, 2013). The biobutanol company Gevo is also

focusing on converting some of its alcohol product to jet fuel blendstocks (Gevo, 2013). In July 2012, the

http://www.google.com.br/patents/US4621164


July 2014 176

US Air Force flew a test flight powered with a 50/50 blend of Gevo’s ATJ blendstock and petroleum JP-8

(Gevo, 2013).

The three hybrid biofuel platforms described in this section have a potential role in drop-in biofuel

production. They also demonstrate that opportunities exist to create effective synergies between

thermo-catalytic and bio-catalytic processes. The performance of these hybrid platform processes again

demonstrate the repeating theme of this report that significant trade-offs exist between drop-in

intermediate or product yields and process inputs. As with the thermochemical and biochemical

platform processes, there is a need to either sacrifice yield on feedstock or consume substantial

amounts of hydrogen to produce highly reduced carbon molecules that are well suited for drop-in

biofuel production from biomass derived feedstocks whether these be sugars, synthesis gases or

alcohols. However, direct comparisons of these processes’ potential economic feasibilities are not

currently possible due to a lack of detailed public domain information on their techno-economic

performance. It is apparent that each hybrid platform has unique advantages such as fuller utilization of

feedstock carbon in syngas fermentations, shorter reaction times in sugar catalytic upgrading or lower

technical risk in alcohol-to-jet technologies. However, as discussed above, each approach also has

significant challenges to overcome before they can be economically scaled up for drop-in biofuel

applications. Encouragingly, INEOS Bio’s and LanzaTech’s syngas fermentation technologies are currently

operating at relatively large demonstration scales and appear to be making good progress in their

respective commercialization strategies. While the other hybrid routes are also progressing, they are still

mostly at the R&D and pilot stages and the timing of larger scale demonstrations for drop-in biofuels

applications remains unclear.


July 2014 177

CHAPTER 6: CONCLUSIONS AND RECOMMENDATIONS

This report defines drop-in biofuels as “liquid bio-hydrocarbons that are functionally equivalent to

petroleum fuels and are fully compatible with existing petroleum infrastructure”. Although conventional

biofuels such as ethanol and biodiesel have been produced and utilized at significant scales for many

years, they are not fully compatible with existing combustion engines or the overall liquid transportation

fuel refining and distribution network. Tremendous entrepreneurial activity to develop and

commercialize drop-in biofuels from aquatic and terrestrial feedstocks has taken place over the past

several years. However, despite these efforts, drop-in biofuels represent only a small percentage

(around 2%) of global biofuel markets. The predominant drop-in fuels produced today are being made

using oleochemical platform routes to convert relatively highly reduced renewable feedstocks such as

palm oil, crops such as rape seed (Canola), used cooking oils, tallow, etc., to fully saturated products.

Much more highly oxygenated feedstocks – lignocellulosic biomass, sugars and ethanol – are being

assessed for thermochemical, biochemical and hybrid platform routes to drop-in biofuels. Most

thermochemical based processes intend to use some form of biomass feedstock and current production

technologies are mainly proven at pilot or demonstration scales. However, some larger scale

thermochemical based processes exist, such as KiOR’s 49 MLPY (13 MGPY) fast pyrolysis facility in

Mississippi, USA, (although success is now looking more problematic due to KiOR’s recently announced

$350 million net loss for 2013). Sugars (also from starch or biomass) are the primary feedstock for most

biochemical based approaches to making drop-in biofuels. The higher energy products (Heff/C ≥ 1)

typically produced by this route can be readily upgraded to drop-in biofuels, although they are also well

suited for applications in the rapidly growing bio-based chemicals market. Regardless of the production

platform, the major interconnected parameters that are influencing the rate of commercialization of

drop-in biofuels are capital and operating expenditures, process yields and productivities, and feedstock

sourcing. As described in greater detail in the main chapters of this report, each platform approach has a

different degree of sensitivity to each of these parameters.

Oleochemical platform processes exhibit high yields, require relatively low capital and hydrogen inputs

(due to the use of relatively high Heff/C feedstocks compared to lignocellulosic or sugar feedstocks) and

are currently at a relatively high level of technological maturity. This is reflected by oleochemical based

drop-in biofuels already being produced at large scale and approved for use by ASTM since July 2011.

However, sourcing large quantities of feedstock (lipids) at a low enough cost to result in profitable drop-


July 2014 178

in biofuel production remains challenging and is a major constraint limiting the expansion of this

platform.

Thermochemical platform pyrolysis based processes require significant hydrogen inputs and specialized

heterogeneous hydroprocessing catalysts that have not, as yet, been proven at scale. Although pyrolysis

oils should be relatively inexpensive to make (about USD $10/GJ based on a USD $83/t feedstock,

according to Bridgwater, 2012; Wright et al. 2010 and Henrich et al., 2009), they contain up to 40 wt%

oxygen and are thus costly to hydroprocess and upgrade to highly saturated drop-in biofuel products.

Upgrading of pyrolysis oils to transportation fuel blendstocks has been estimated to account for about

two thirds of capital costs and one third of operating expenses (Jones et al., 2009). Given current

technology status and market trends, it is unlikely in the near to midterm that hydrogen required for

upgrading will be generated from biomass feedstocks at the expense of fuel yields and more likely it will

be sourced from fossil natural gas. It has been estimated that hydrogen generation capacity in US

refineries would have to be tripled to meet the US RFS targets of 15 billion advanced biofuels by 2022

with biomass pyrolysis fuels. Due to the heterogeneous nature of bio-oils it has been suggested that a

two-stage hydrotreatment would be a more cost effective approach to bio-oil upgrading (Elliot, 2007;

Jones et al., 2009). The first step would stabilize the bio-oil by selectively hydrotreating its most reactive

(unstable) organic species and, after stabilization a second step would be used to complete

hydrotreatment. This two-step approach lends itself to a spoke and hub model where the first

stabilization step is performed at many smaller scale facilities located near feedstock supplies and the

second final upgrading step is performed at a few larger centralized facilities – perhaps petroleum

refineries – where greater economies of scale are possible and existing capital assets, if available, can be

used. Once hydrogen sourcing and catalyst issues are resolved, the pyrolysis platform holds great

potential since it can effectively utilize a range of biomass feedstocks and has relatively low capital costs

(compared to gasification), particularly if it can be developed to leverage the substantial existing

petroleum refining infrastructure.

Most proposed pyrolysis technologies that have been commercially pursued use reactor designs not

dissimilar to fluidized catalytic crackers (FCCs) located at modern oil refineries. Both the FCCs and

hydroprocessing units within an oil refinery could potentially be used to upgrade bio-oils or bio-oil-

petroleum blends. Currently, a considerable amount of the FCC capacity in the US is not fully utilised,

creating an opportunity to leverage these “stranded” assets. However, using these FCC facilities will be

challenging as conventional FCC processing takes place in the absence of hydrogen. The processing of


July 2014 179

lower Heff/C pyrolytic oils is likely to exacerbate yield losses and coking in FCC units. Current petroleum

catalysts are incompatible with highly oxygenated biomass derived compounds and biomass feedstocks

increase the risk of contaminating downstream refinery units with oxygenated and inorganic species

(originating from biofeeds). There is also a large disparity in scale between pyrolysis and refinery FCCs.

Encouragingly, despite these challenges, partially upgraded (20% oxygen) pyrolysis oils have been

successfully inserted into an FCC, demonstrating the potential viability of this type of approach and

resulting in significant techno-economic benefits (Solantausta, 2011). These include better use of

existing refining equipment, easier access to hydrogen as well as the ability to use any excess coke and

gas formed as combustion fuels for the process. With current technology, pyrolysis oil hydroprocessing

in oil refineries requires the bio-oil oxygen content to be lower than about 5 wt%. Alternatively,

blending higher oxygen content bio-oils with petroleum liquids is another way of maintaining

oxygenated species at acceptable levels for processing in oil refinery catalytic reactors (Huber et al.,

2007; Bunting et al., 2009; Jones et al., 2009).

The greatest challenge for gasification based production routes to drop-in biofuels is the likelihood that

high capital expenditures will be required to meet the capital equipment economies of scale needed to

achieve cost competitive economics. A related challenge is the logistics and costs of sourcing sufficient

amounts of biomass feedstock required to operate a large facility. The capital and operating expenses of

gasification platform processes are estimated by techno-economic analyses to be about double those

for pyrolysis (Anex et al., 2010), however the technology risk for gasification is thought to be

significantly lower than for pyrolysis since biomass gasification is already practiced commercially (about

500 MW capacity globally) and upgrading of fossil fuel based syngas to liquid fuels is practiced at scales

that exceed 20 billion L/year. Large scale fossil fuels-to-liquids facilities include Sasol’s coal-to-liquid

facility in South Africa that has been operating since the 1980s and Shell’s recently built Pearl GtL

natural gas-to-liquid facility in Qatar. Although these commercial plants have demonstrated the

industrial robustness of FT catalysts, they are each about 100 times larger than current “large scale”

biofuel facilities. Moreover, the processing of more highly oxygenated biomass derived syngas is

expected to result in considerably more challenges than coal or natural gas feedstocks. Biomass syngas

has a lower energy density (volumetrically up to 4 times lower), a lower H2/CO ratio, and orders of

magnitude higher concentrations of problematic contaminants such as tars (if not sufficiently mitigated

in the conversion process) than fossil fuel derived syngas. To deal with these shortcomings, biomass

syngas is often extensively cleaned and hydrogen is added, either by oxidizing carbon monoxide (via the


July 2014 180

water-gas shift reaction) or by inserting externally supplied hydrogen (e.g., as practiced by Sundrop

Biofuels). These processes add to the cost of gasification-derived biofuels. The syngas can be upgraded

to liquid fuels via several routes such as mixed alcohol synthesis and then catalytic upgrading to drop-in

biofuels (e.g., the MTG or ATJ processes), DME synthesis and then further upgrading to drop-in biofuels,

or direct production to drop-in biofuels via FT synthesis. The efficiency of these processes can be roughly

ranked as alcohols > DME > MTG > FT, and this trend generally follows the oxygen content and Heff/C

ratios of the products generated by these thermochemical syngas upgrading processes. The use of

biomass pellets or pyrolysis oils might help gasification platforms deal with some feedstock logistic

challenges encountered at a large scale. Alternatively, companies such as Velocys claim that small scale

gasification systems can be built and operated commercially, however their economic feasibility has yet

to be proven.

It has been suggested that one of the advantages of biochemical platform technologies for producing

drop-in biofuel intermediates is the potential for new processes to be “bolted on” to existing sugar and

starch mills to leverage existing biorefining assets and thereby reduce capital and operating costs

(Amyris, 2013a; Gevo, 2013). The metabolic pathways used to convert oxygen-rich sugar feedstocks

(Heff/C = 0) to highly reduced products such as fatty acids and isoprenoids (Heff/C = 1.8 – 2) for drop-in

biofuel production are, however, energy- and carbon-intensive compared to conventional sugar

fermentation. For example, fatty acid biosynthesis requires a significant portion of the feedstock carbon

to be consumed to produce the energy required to drive fatty acid elongation. Each elongation cycle

requires one reducing NADPH molecule to drive biosynthesis, such that one molecule of a C16 fatty acid

(i.e., palmitic acid) requires 7 repetition cycles and thus 7 NADPH molecules, which are usually

generated by oxidizing feedstock carbon at the expense of product yield. Even operating aerobically,

reported biological routes to convert sugars to drop-in intermediates have so far only achieved

volumetric productivities that are an order of magnitude lower than anaerobic sugar-to-ethanol

fermentations (Table 4-2). However, with fewer processing steps and lower NADPH requirements (or

hydrogen consumption), biochemical platform routes are already well suited to make oxygenated

products such as carboxylic acids, alcohols and polyols that can generate high revenues in the rapidly

growing bio-based chemicals markets. Less oxygenated microbial metabolites with potential to be drop-

in biofuel intermediates such as Amyris’s farnesene and Gevo or Butamax’s butanol are already being

sold in the value added chemicals and cosmetics markets. Thus, market competition for biochemical

drop-in products is expected to remain significant and grow in the near to mid-term as markets for


July 2014 181

biobased chemicals continue to expand. Estimates are that these markets will reach 10-50 million

tonnes per year by 2020, equal to the current market size for biofuels (Lux Research, 2010; Higson,

2011; Bomgardner, 2012).

Several hybrid platforms for drop-in biofuel are also under active commercial development, including

catalytic aqueous phase reforming (APR) (e.g., Virent), alcohol-to-jet (ATJ) (e.g., Cobalt, Gevo), acid-to-

alcohol (e.g., Zeachem) and syngas fermentation (e.g., INEOS Bio, LanzaTech). Based on the limited

information in the public domain, each of these approaches appears to have distinct advantages but

also significant challenges. For example, the APR process can produce drop-in fuel blendstocks but relies

on new catalysts that have not, as yet, been proven at industrial scale or over prolonged operations with

highly oxidized (hydrogen-deficient) renewable feeds. Similarly, syngas fermentation offers the potential

to convert higher proportions of feedstock carbon to product but the limited solubility of the CO gas in

the aqueous solutions makes scale up difficult, and the need to cool syngas from over 700 °C to below

50 °C for fermentation hinders process thermal efficiency. However, two hybrid platform companies

focused on syngas fermentation technology -- INEOS Bio and LanzaTech – have built and are operating

large demonstration scale or small commercial scale facilities. Although ATJ and acid-to-alcohol

technologies are technically proven, the would-be drop-in intermediates or their feedstocks, such as

ethanol, acetic acid, ethyl acetate, etc., are typically more valuable (on an energy basis) as chemicals

than as jet or alcohol fuel products. However, jet fuels from alcohols are expected to acquire ASTM

certification in early 2014, which raises the possibility that ATJ drop-in jet fuel could become

commercially attractive in countries such as Brazil and the US where ethanol is available at relatively low

cost.

Hydroprocessing in one form or another is a final processing step shared by almost all drop-in biofuel

platforms. This is the step where FT liquids, hydrotreated pyrolysis oils, HEFA lipid feedstocks and

partially oxygenated or unsaturated biochemical platform drop-in intermediates are upgraded to

gasoline, diesel or jet fuel blendstocks. Therefore, a large as yet mostly unrealized opportunity exists to

better leverage existing oil refinery infrastructure by employing hydroprocessing units to upgrade drop-

in biofuel intermediates. However, recent trials where vegetable oils were co-processed with petroleum

in hydrocrackers showed that current catalysts have limited ability to effectively process feeds

containing moderate levels of oxygenated species. This relatively simple vegetable oil insertion into an

oil refinery’s hydrocracker is, so far, the only reported industrial scale example of biofeed processing in a

petroleum refinery.


July 2014 182

Due to the increased processing and resource requirements (e.g., hydrogen and catalysts) needed to

make drop-in biofuels as compared to conventional biofuels, large scale production of cost-competitive

drop-in biofuels is not expected to occur in the near to midterm. Rather, dedicated policies to promote

development and commercialization of these fuels will be needed before they become significant

contributors to global biofuels production. Currently, no policies (e.g., tax breaks, subsidies etc.)

differentiate new, more fungible and infrastructure ready drop-in type biofuels from less infrastructure

compatible oxygenated biofuels. While some mandates distinguish between conventional and advanced

biofuels in terms of their greenhouse gas emissions reduction potential (e.g., the US RFS), there are

currently no such incentives for drop-in biofuels. While oleochemical platforms are commercial today,

their future growth is constrained by the cost and availability of feedstocks. Although thermochemical

based processes have considerable potential ongoing R&D is needed to resolve scale and catalyst issues.

In the short term, biochemical platform routes to drop-in fuels face stiff market competition from

biobased chemicals markets and as a result will likely focus on the production of higher value chemicals

while building the experience and capacity needed to be able to scale up drop-in biofuel production

when the demands for chemicals plateau. Several hybrid routes to jet fuel from alcohols are progressing

to support specific market opportunities but here again cost competitiveness remains a challenge. Thus,

while tremendous technical progress has been made in developing and improving the various routes to

drop-in fuels, supportive policies directed specifically towards the further development of drop-in

biofuels are likely to be needed to ensure their future commercial success.


July 2014 183

References

Aatola, H., Larmi, M., Sarjovaara, T., & Mikkonen, S. (2008). Hydrotreated Vegetable Oil (HVO) as a Renewable Diesel Fuel: Trade-off between NOx, Particulate Emission, and Fuel Consumption of a Heavy Duty Engine. SAE Technical Paper 2008-01-2500. SAE Technical Papers (p. 12). doi:10.4271/2008-01-2500

Akhtar, J., & Amin, N. A. S. (2011). A review on process conditions for optimum bio-oil yield in hydrothermal liquefaction of biomass. Renewable and Sustainable Energy Reviews, 15(3), 1615–1624. doi:10.1016/j.rser.2010.11.054

Alberta Government. (2007, June 20). Facts and Statistics. Alberta Energy. Retrieved June 10, 2013, from http://www.energy.alberta.ca/oilsands/791.asp

Alexander, D. (2012, July). U.S. Air Force tests biofuel at $59 per gallon. Reuters News. Retrieved from http://www.reuters.com/article/2012/07/15/us-usa-military-biofuels-idUSBRE86E01N20120715

Amyris. (2013a). Amyris - official website. various sections. Retrieved August 23, 2013, from www.amyris.com

Amyris. (2013b). Amyris - Products - Diesel. Press releases. Retrieved August 28, 2013, from http://www.amyris.com/Content/Detail.aspx?ReleaseID=166&NewsAreaID=21&ClientID=1

Anellotech. (2013). Anellotech, Inc. Retrieved August 07, 2013, from http://www.anellotech.com/

ATAG. (2011). Where are we with biofuels - Presentation (p. 19). Montreal, Quebec, Canada. Retrieved from http://www.icao.int/Meetings/EnvironmentalWorkshops/Documents/2011-SUSTAF/18_Steele.pdf

ATAG. (2012). Powering the future of flight, March 2012 (p. 20). Retrieved from http://www.atag.org/component/downloads/downloads/152.html

Aziz, N. A., & Dato’ Leon, K. M. (2012). Malaysia’s biomass potential. Biomass SP, BE Sustainable, (April), 35–39. Retrieved from http://issuu.com/besustainablemagazine/docs/besustainablemagazine-issue_0

Babu, S. (2005). Biomass gasification for hydrogen production–process description and research needs. In Technology Report from IEA Bioenergy Task 33 for ExCo56 (p. 11). Retrieved from http://www.ieabioenergy.com/wp-content/uploads/2013/10/IEA-Bioenergy-Update-35-Task-33-Technology-Report.pdf

Bai, F. W., Anderson, W. a, & Moo-Young, M. (2008). Ethanol fermentation technologies from sugar and starch feedstocks. Biotechnology Advances, 26(1), 89–105. doi:10.1016/j.biotechadv.2007.09.002

Bain, R. L. (1992). Material and Energy Balances for Methanol from Biomass Using Biomass Gasifiers (p. 136). Golden, CO, USA. Retrieved from http://www.nrel.gov/docs/legosti/old/17098.pdf


July 2014 184

Bain, R. L., & Broer, K. (2011). Gasification. In R. C. Brown (Ed.), Thermochemical Processing of Biomass: Conversion into Fuels, Chemicals and Power (1st ed., pp. 47–77). Chichester UK: Wiley-VCH.

Balat, M., Balat, M., Kırtay, E., & Balat, H. (2009). Main routes for the thermo-conversion of biomass into fuels and chemicals. Part 1: Pyrolysis systems. Energy Conversion and Management, 50(12), 3147–3157. doi:10.1016/j.enconman.2009.08.014

Bauen, A., Howes, J., Bertuccioli, L., & Chudziak, C. (2009). Review of the Potential for Biofuels in Aviation. E4Tech. Retrieved from http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:Review+of+the+potential+for+biofuels+in+aviation#0

Bauldreay, J. M., Blommel, P. G., Cortright, R. D., & Price, R. J. (2010). LIQUID FUEL COMPOSITIONS. Retrieved from http://www.patentlens.net/patentlens/patents.html?patnums=US_2010_0218417_A1&language=en&query=(US_2010_0218417_A1 in publication_number)&stemming=true&returnTo=structured.html?query=US+2010/0218417+A1

Bayar, T. (2013, March). Sweden’s Bioenergy Success Story. Renewable Energy World. Retrieved from http://www.renewableenergyworld.com/rea/news/article/2013/03/swedens-bioenergy-success-story

Bezergianni, S., Kalogianni, A., & Vasalos, I. a. (2009). Hydrocracking of vacuum gas oil-vegetable oil mixtures for biofuels production. Bioresource Technology, 100(12), 3036–42. doi:10.1016/j.biortech.2009.01.018

Binder, M., & Woods, L. (2009). Final Report Comparative Life Cycle Assessment Ingeo TM biopolymer , PET , and PP by (p. 61). Boston, MA, USA.

BioDME. (2013). BioDME. Press releases. Retrieved August 28, 2013, from http://www.biodme.eu/

Biofuels-Journal. (2010). Butamax Opens Biobutanol Technology Laboratory in Brazil. Advanced Biofuels. Retrieved from Butamax Opens Biobutanol Technology Laboratory in Brazil

Blades, T., Rudloff, M., & Schulze, O. (2005). Sustainable SunFuel from CHOREN ’ s Carbo-V ® Process. In ISAF XV, San Diego, September 2005 (pp. 1–6). San Diego CA. Retrieved from http://www.eri.ucr.edu/ISAFXVCD/ISAFXVAF/SSFCCVP.pdf

Blommel, P. G., & Cortright, R. D. (2008). Production of Conventional Liquid Fuels from Sugars.

Bomgardner, M. M. (2012, September). Biobased summer. C&EN, 10–13.

Bradley, D. (2009). 2nd Generation Biofuels and Trade An exploratory study (pp. 1–55). Retrieved from http://www.bioenergytrade.org/downloads/2ndgenerationbiofuelsfinaldec142009.pdf

Bridgwater, A. V. (2012). Review of fast pyrolysis of biomass and product upgrading. Biomass and Bioenergy, 38, 68–94. doi:10.1016/j.biombioe.2011.01.048


July 2014 185

Brown, R. C. (2011). Thermochemical Processing of Biomass: Conversion into Fuels, Chemicals and Power. (R. C. Brown, Ed.) (1st ed., p. 330). Hoboken, NJ, USA: Wiley-VCH. Retrieved from http://onlinelibrary.wiley.com/book/10.1002/9781119990840

Brown, T. R., Tilakaratne, R., Brown, R. C., & Hu, G. (2012). Techno-economic analysis of biomass to transportation fuels and electricity via fast pyrolysis and hydroprocessing. Fuel. doi:10.1016/j.fuel.2012.11.029

Bryan, P. (2011). Turning Waste Into Fuel: How the INEOS Biorefinery Is Changing the Clean Energy Game. News from Biomass Program, Office of Energy Efficiency & Renewable Energy. Retrieved from http://energy.gov/articles/turning-waste-fuel-how-ineos-biorefinery-changing-clean-energy-game

BTG. (2012). Biomass Technology Group - official website. Technologies/Fast pyrolysis. Retrieved August 07, 2013, from http://www.btgworld.com/en/rtd/technologies/fast-pyrolysis

Bui, V. N., Toussaint, G., Laurenti, D., Mirodatos, C., & Geantet, C. (2009). Co-processing of pyrolisis bio oils and gas oil for new generation of bio-fuels: Hydrodeoxygenation of guaïacol and SRGO mixed feed. Catalysis Today, 143(1-2), 172–178. doi:10.1016/j.cattod.2008.11.024

Bunting, B., Bunce, M., Brone, T., & Storey, J. (2010). Fungible and compatible biofuels: literature search , summary , and recommendations (p. 112). Oak Ridge, TN, USA. Retrieved from https://www.bioenergykdf.net/sites/default/files/Bruce_Bunting/20110512135206-0.pdf

Butler, E., Devlin, G., Meier, D., & McDonnell, K. (2011). A review of recent laboratory research and commercial developments in fast pyrolysis and upgrading. Renewable and Sustainable Energy Reviews, 15(8), 4171–4186. doi:10.1016/j.rser.2011.07.035

Byogy. (2013). Byogy - official website. various sections. Retrieved August 28, 2013, from www.byogy.com

Campbell, J. E., Lobell, D. B., & Field, C. B. (2009). Greater transportation energy and GHG offsets from bioelectricity than ethanol. Science (New York, N.Y.), 324(5930), 1055–7. doi:10.1126/science.1168885

Carpenter, D. L., Bain, R. L., Davis, R. E., Dutta, A., Feik, C. J., Gaston, K. R., … Nimlos, M. R. (2010). Pilot-Scale Gasification of Corn Stover, Switchgrass, Wheat Straw, and Wood: 1. Parametric Study and Comparison with Literature. Industrial & Engineering Chemistry Research, 49(4), 1859–1871. doi:10.1021/ie900595m

CEN. (2014). European Committee for Standardization, technical committee on gaseous and liquid fuels, lubricants and related products of petroleum, syntehtic and biological origin, list of working groups. official website. Retrieved from http://www.cen.eu/cen/Sectors/TechnicalCommitteesWorkshops/CENTechnicalCommittees/Pages/TCStruc.aspx?param=6003&title=Gaseous and liquid fuels, lubricants and related products of petroleum, synthetic and biological origin.


July 2014 186

Centeno, A., Laurent, A., & Delmon, B. (1995). Influence of the Support of CoMo Sulfide Catalysts and of the Addition of Potassium and Platinum on the Catalytic Performances for the Hydrodeoxygenation of Carbonyl, Carboxyl, and Guaiacol-Type Molecules. Journal of Catalysis, 154(2), 288–298. doi:10.1006/jcat.1995.1170

Chemrec. (2009). A brand new renewable engine fuel from the forest. Chemrec News. Retrieved August 30, 2013, from http://www.chemrec.se/SunPine_producing_tall_oil_diesel.aspx

Connor, D. O. (2013). Advanced Biofuels – GHG Emissions and Energy Balances. vancouver. Retrieved from http://task39.org/files/2013/05/Energy-and-GHG-Emissions-IEA-Bioenergy-T39-Report-May-2013-rev.pdf

ConocoPhillips. (2010). Renewable Fuels: A Refiner’s Perspective. Retrieved May 21, 2013, from http://www.topsoe.com/research/Topsoe_Catalysis_Forum.aspx~/media/PDF files/Topsoe_Catalysis_Forum/2010/tushar.ashx

Conocophillips. (2013). ConocoPhillips - official website. company reports - the whitegate story. Retrieved August 27, 2013, from http://www.conocophillips.com/EN/about/company_reports/spirit_mag/Pages/whitegate_story.aspx

Corma, A., Huber, G., Sauvanaud, L., & Oconnor, P. (2007). Processing biomass-derived oxygenates in the oil refinery: Catalytic cracking (FCC) reaction pathways and role of catalyst. Journal of Catalysis, 247(2), 307–327. doi:10.1016/j.jcat.2007.01.023

Cortright, R., Davda, R., & Dumesic, J. (2002). Hydrogen from catalytic reforming of biomass-derived hydrocarbons in liquid water. Nature, 418(August), 54–57. Retrieved from http://www.nature.com/nature/journal/v418/n6901/abs/nature01009.html

COWI. (2013). Black Liquor to Methanol or DME Technical information - Black Liquor to Methanol or DME. Retrieved from http://www.cowi.se/menu/projekt/Industri-och-Energi/Documents/Black Liquor to Methanol or DME Chemrec.pdf

Craig, K. R., & Mann, M. K. (1996). Cost and Performance Analysis of Biomass-Based Integrated Gasification Combined-Cycle ( BIGCC ) Power Systems (p. 70). Golden, CO, USA. Retrieved from http://www.nrel.gov/docs/legosti/fy97/21657.pdf

Czernik, S., & Bridgwater, A. (2004). Overview of applications of biomass fast pyrolysis oil. Energy & Fuels, (12), 590–598. Retrieved from http://pubs.acs.org/doi/abs/10.1021/ef034067u

Dahmen, N., Henrich, E., Dinjus, E., & Weirich, F. (2012). The bioliq® bioslurry gasification process for the production of biosynfuels, organic chemicals, and energy. Energy, Sustainability and Society, 2(1), 3. doi:10.1186/2192-0567-2-3

Daniell, J., Köpke, M., & Simpson, S. (2012). Commercial Biomass Syngas Fermentation. Energies (Vol. 5, pp. 5372–5417). doi:10.3390/en5125372


July 2014 187

Darzins, A., Pienkos, P., & Edye, L. (2010). Current Status and Potential for Algal Biofuels Production, (August).

Dayton, D., Turk, B., & Gupta, R. (2011). Syngas cleanup, conditioning and utilization. In R. C. Brown (Ed.), Thermochemical Processing of Biomass: Conversion into Fuels, Chemicals and Power (1st ed., pp. 78–123). Chichester UK: Wiley-VCH.

De Almeida, E. F., Bomtempo, J. V., & de Souza e Silva, C. M. (2008). The performance of Brazilian biofuels: An economic, environmental and social analysis. In OECD and International Transport Forum, Biofuels: Linking Support to Performance. OECD Publishing. doi:10.1787/9789282101803-6-en

De Miguel Mercader, F., Groeneveld, M. J., Kersten, S. R. a., Geantet, C., Toussaint, G., Way, N. W. J., … Hogendoorn, K. J. a. (2011). Hydrodeoxygenation of pyrolysis oil fractions: process understanding and quality assessment through co-processing in refinery units. Energy & Environmental Science, 4(3), 985. doi:10.1039/c0ee00523a

De Miguel Mercader, F., Groeneveld, M. J., Kersten, S. R. a., Way, N. W. J., Schaverien, C. J., & Hogendoorn, J. a. (2010). Production of advanced biofuels: Co-processing of upgraded pyrolysis oil in standard refinery units. Applied Catalysis B: Environmental, 96(1-2), 57–66. doi:10.1016/j.apcatb.2010.01.033

De Miguel Mercader, F., Hogendoorn, K. J. a., Venderbosch, R., Kersten, S. R. a., & Groeneveld, M. J. (2010). Pyrolysis oil upgrading by high pressure thermal treatment. Fuel, 89(10), 2829–2837.

Diamond Green Diesel. (2013). Diamond Green Diesel - official website. Homepage. Retrieved July 28, 2013, from http://www.diamondgreendiesel.com/Pages/default.aspx

Diebold, J., & Czernik, S. (1997). Additives to lower and stabilize the viscosity of pyrolysis oils during storage. Energy & Fuels, 0624(10), 1081–1091. Retrieved from http://pubs.acs.org/doi/abs/10.1021/ef9700339

Diebold, J. P. (2000). A Review of the Chemical and Physical Mechanisms of the Storage Stability of Fast Pyrolysis Bio-Oils (p. 59). Golden, CO, USA. Retrieved from http://gisceu.net/PDF/U918.pdf

DOE. (2011). DOE Office of Biomass Program Review. Retrieved July 27, 2013, from http://www.obpreview2011.govtools.us/thermochem/

Drift, A. van der, Boerrigter, H., Coda, B., Cieplik, M. K., & Hemmes, K. (2004). Entrained flow gasification of biomass: Ash behaviour, feeding issues, and system analyses (p. 58). Retrieved from http://www.ecn.nl/docs/library/report/2004/c04039.pdf

Drift, A. Van Der, Doorn, J. Van, & Vermeulen, J. W. (2001). Ten residual biomass fuels for circulating uidized-bed gasiÿcation, 20, 45–56.

Dundon, C. A., Aristidou, A. A., Hawkins, A., Lies, D., & Albert, L. H. (2011). Methods of increasing dihydroxy acid dehydratase activity to improve production of fuels, chemicals, and amino acids.


July 2014 188

USA. Retrieved from http://www.patentlens.net/patentlens/patents.html?patnums=US_8017376&language=en&query=(US_8017376 in publication_number)&stemming=true&returnTo=structured.html?query=(US_8017376+in+publication_number)&stemming=true&collections=US_B,EP_B,AU_B,US_A,WO_A,AU_A&language=en&pageLength=10&fields=publication_number,title,abstract,inventor,applicant

Dynamic Fuels. (2013). Dynamic Fuels - official website. About. Retrieved August 28, 2013, from http://www.dynamicfuelsllc.com/about.aspx

Dynamotive. (2009). Dynamotive | Dynamotive Upgrades BioOil from Multiple Biomass Sources to a Uniform Hydrocarbon Product. Provides Update on Upgraded BioOil® Corroborating Analysis. Retrieved August 06, 2013, from http://www.dynamotive.com/2009/10/28/dynamotive-upgrades-biooil-from-multiple-biomass-sources-to-a-uniform-hydrocarbon-product-provides-update-on-upgraded-biooil-corroborating-analysis/

E4tech. (2009). Review of Technologies for Gasification of Biomass and Wastes Final report (p. 130). London. Retrieved from C:\Users\Sergios\Dropbox\NEW DROP IN\CH 3 THERMO\(http:\wiki.gekgasifier.com\f\Review+of+Biomass+Gasification+Technologies.NNFCC.Jun09.pdf

EBTP. (2013). Euorpean Biofuels Technology Platform - official website. Biomass to Liquids. Retrieved August 28, 2013, from http://www.biofuelstp.eu/btl.html

ECN. (2008). MILENA gasification process (p. 2). Retrieved from https://www.ecn.nl/fileadmin/ecn/units/bio/Leaflets/b-08-021_MILENA.pdf

Egeberg, R. G., Michaelsen, N. H., & Skyum, L. (2010). Novel hydrotreating technology for production of green diesel (p. 21). Lyngby, Denmark. Retrieved from http://www.topsoe.com/business_areas/refining/~/media/PDF files/Refining/novel_hydrotreating_technology_for_production_of_green_diesel.ashx

EIA. (2009). Annual Energy Outlook 2009 with projections to 2030 (p. 230). Washington, DC, USA. Retrieved from http://www.eia.gov/oiaf/aeo/pdf/0383(2009).pdf

EIA. (2013). Annual Energy Outlook 2013: Early Release Overview (p. 16). Washington, DC, USA. Retrieved from http://www.eia.gov/forecasts/aeo/er/pdf/0383er(2013).pdf

Elliott, D. C. (2007). Historical Developments in Hydroprocessing Bio-oils. Energy & Fuels, 21(3), 1792–1815. doi:10.1021/ef070044u

Emerald Biofuels. (2013). Emerald Biofuels - official website. Projects. Retrieved August 29, 2013, from http://www.emeraldbiofuels.com/projects.php

Empyro. (2013). Pyrolysis oil - Empyro - energy & materials from pyrolysis. Retrieved August 28, 2013, from http://empyroproject.eu/index.php?id=12&


July 2014 189

Ensyn. (2013a). Ensyn - official website. projects/renfrew-ontario. Retrieved June 25, 2013, from http://www.ensyn.com/projects/renfrew-ontario/

Ensyn. (2013b). Ensyn I RTP Renewable Liquid Fuels from Residual Biomass. Retrieved August 07, 2013, from http://www.ensyn.com/

Essent. (2013). ISSUU - Natural Power: Essent’s vision for a bio-based economy. ISSUU. Retrieved August 28, 2013, from http://issuu.com/essentnl/docs/bio-based-economy

European Commission. (2001). Ambient air pollution by Polycyclic Aromatic Hydrocarbons ( PAH ). Retrieved from https://www.google.ca/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&ved=0CDEQFjAA&url=http%3A%2F%2Fec.europa.eu%2Fenvironment%2Fair%2Fpdf%2Fpp_pah.pdf&ei=Fn3RUpPzMcb6oASKvYCoDg&usg=AFQjCNHuPetBH5hlhEYcEtGr3YFyyt9tLw&sig2=jVDOglH4ikdA_tMPBQr63Q&bvm=bv.59026428,d.cGU

Evanko, W. A., Eyal, A. M., Glassner, D. A., Miao, F., Aristidou, A. A., Evans, K., … Hawkins, A. (2012). Recovery of higher alcohols from dilute aqueous solutions. USA. Retrieved from http://gevo.com/assets/patents/US8283505.pdf

Ezeji, T., Milne, C., Price, N. D., & Blaschek, H. P. (2010). Achievements and perspectives to overcome the poor solvent resistance in acetone and butanol-producing microorganisms. Applied Microbiology and Biotechnology, 85(6), 1697–712. doi:10.1007/s00253-009-2390-0

Feldman, R. M., Gunawardena, U., Urano, J., Meinhold, P., Aristidou, A. A., Dundon, C. A., & Smith, C. (2008). Yeast organism producing isobutanol at high yield. USA. Retrieved from http://www.patentlens.net/patentlens/patents.html?patnums=US_8017375&language=en&query=(US_8017375 in publication_number)&stemming=true&returnTo=structured.html?query=(8,017,375+in+publication_number)&stemming=true&collections=US_B,EP_B,AU_B,US_A,WO_A,AU_A&language=en&pageLength=10&fields=publication_number,title,abstract,inventor,applicant#tab_0

Ferrari, M., Bosmans, S., Maggi, R., Delmon, B., & Grange, P. (2001). CoMo/carbon hydrodeoxygenation catalysts: influence of the hydrogen activity of a CoMo/carbon catalyst. Catalysis Today, 65(2-4), 257–264.

Finkbeiner, M. (2013). Indirect land use change (iLUC) within life cycle assessment (LCA) – Scientific robustness and consistency with international standards (p. 62). Retrieved from http://www.ufop.de/files/5213/6853/0753/RZ_VDB_0030_Studie_ENG_Komplett.pdf

Fisk, C. a., Morgan, T., Ji, Y., Crocker, M., Crofcheck, C., & Lewis, S. a. (2009). Bio-oil upgrading over platinum catalysts using in situ generated hydrogen. Applied Catalysis A: General, 358(2), 150–156. doi:10.1016/j.apcata.2009.02.006

FOE. (2012). Biokerosene : Take-off in the wrong direction (p. 28). Retrieved from https://www.milieudefensie.nl/publicaties/rapporten/take-off-in-the-wrong-direction


July 2014 190

Fogassy, G., Thegarid, N., Toussaint, G., van Veen, A. C., Schuurman, Y., & Mirodatos, C. (2010). Biomass derived feedstock co-processing with vacuum gas oil for second-generation fuel production in FCC units. Applied Catalysis B: Environmental, 96(3-4), 476–485. doi:10.1016/j.apcatb.2010.03.008

Food and Agriculture Organization (FAO). (2012). Current world fertilizer trends and outlook to 2016. Rome. Retrieved from ftp://ftp.fao.org/ag/agp/docs/cwfto16.pdf

Forsberg, C. W. (2005). Hydrogen Markets : Implications for Hydrogen Production Technologies (p. 13). Oak Ridge, TN, USA. Retrieved from http://www.ornl.gov/~webworks/cppr/y2001/pres/122902.pdf

Forsberg, C. W. (2009). Meeting U.S. liquid transport fuel needs with a nuclear hydrogen biomass system. International Journal of Hydrogen Energy, 34(9), 4227–4236. doi:10.1016/j.ijhydene.2008.07.110

Fortum. (2013a). Fortum to start supplying bio-oil to a local energy company. Press releases. Retrieved November 15, 2013, from http://www.fortum.com/en/mediaroom/pages/fortum-to-start-supplying-bio-oil-to-a-local-energy-company.aspx

Fortum. (2013b). Fortum’s bio-oil plant commissioned in Joensuu - first of its kind in the world. Press releases. Retrieved November 30, 2013, from http://www.fortum.com/en/mediaroom/pages/fortums-bio-oil-plant-commissioned-in-joensuu-first-of-its-kind-in-the-world.aspx

Foster Wheeler. (2013). Foster Wheeler - official website. Biomass Gasifiers & Co-firing Systems. Retrieved August 29, 2013, from http://www.fwc.com/What-We-Do/Power/Auxiliary-Equipment/Biomass-Gasifiers.aspx

Foust, T. (2011). National Advanced Biofuels Consortium (NABC) presentation at the 2011 DOE Biomass Program Review (p. 52). Washington DC. Retrieved from http://www.obpreview2011.govtools.us/presenters/public/InsecureDownload.aspx?filename=NABC DOE Peer Review-Foust_FINAL-BC.pdf

FPAC. (2011). The New Face of the Canadian Forest Industry Bio-pathways (p. 12). Retrieved from http://www.fpac.ca/publications/BIOPATHWAYS II web.pdf

Furimsky, E., & Massoth, F. (1999). Deactivation of hydroprocessing catalysts. Catalysis Today, 52. Retrieved from http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:Deactivation+of+hydroprocessing+catalysts#0

Gary, J., Handwerk, G. E., & Kaiser, M. J. (2007). Petroleum Refining and Economics (5th ed., p. 488). Golden, CO, USA: CRC Press.

Gerdes, K., & Skone, T. (2009). An evaluation of the extraction, transport and refining of imported crude oils and the impact on life cycle greenhouse gas emissions (p. 42). Morgantown, WV, USA. Retrieved from


July 2014 191

http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:An+Evaluation+of+the+Extraction+,+Transport+and+Refining+of+Imported+Crude+Oils+and+the+Impact+on+Life+Cycle+Greenhouse+Gas+Emissions#0

Gevo. (2013). Gevo - official website. various sections. Retrieved August 23, 2013, from www.gevo.com

Gilmore, C. K., Barrett, S. R. H., Yim, S. H. L., Murray, T., Kuhn, S. R., Tai, A. P. K., … Waitz, I. A. (2011). Environmental Cost- Benefit Analysis of Ultra Low Sulfur Jet Fuel Environmental Cost-Benefit Analysis of Ultra Low Sulfur Jet Fuel (p. 140). Cambridge, MA. Retrieved from http://partner.mit.edu/projects/environmental-cost-benefit-analysis-ultra-low-sulfur-jet-fuels

Goudrian, F., & Peferoen, D. G. R. (1990). Liquid fuels form biomass via a hydrothermal process. Chemical Engineering Science, 45(8), 2729–2734.

Green Fuel Nordic. (2013). Green Fuel Nordic - official website. various sections. Retrieved August 25, 2013, from www.greenfuelnordic.fi

Gunnarsson, I. (2013). The GoBiGas-project (p. 20). Oslo, Norway. Retrieved from http://nobio.no/upload_dir/pics/Ingemar-Gunnarsson.pdf

Hannon, J. R., Bakker, A., Lynd, L. R., & Wyman, C. W. (2007). Comparing the Scale-Up of Anaerobic and Aerobic Processes. In American Institute of Chemical Engineers, Salt Lake City, Utah, November 8, 2007 (p. 15). Salt Lake City, Utah, USA. Retrieved from http://www.bakker.org/cfm/publications/AIChE-2007-Hannon.pdf

Hayes, D. J. M. (2013). Second-generation biofuels: why they are taking so long. Wiley Interdisciplinary Reviews: Energy and Environment, 2(3), 304–334. doi:10.1002/wene.59

Henrich, E., Dahmen, N., Dinjus, E., & Karlsruhe, F. (2009). Cost estimate for biosynfuel production via biosyncrude gasifi cation. doi:10.1002/bbb

Higson, A. (2011). Biorefining and Bio-based Products Today ’ s Presentation About the NNFCC, (February).

Hileman, J., Ortiz, D., Bartis, J., & Wong, H. (2009). Near-term feasibility of alternative jet fuels (p. 150). Santa Monica, CA. Retrieved from http://stuff.mit.edu:8001/afs/athena.mit.edu/dept/aeroastro/partner/reports/proj17/altfuelfeasrpt.pdf

Hofvendahl, K., & Hahn-Hägerdal, B. (2000). Factors affecting the fermentative lactic acid production from renewable resources. Enzyme and Microbial Technology, 26(2-4), 87–107. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10689064

Holladay, J. D., Hu, J., King, D. L., & Wang, Y. (2009). An overview of hydrogen production technologies. Catalysis Today, 139(4), 244–260. doi:10.1016/j.cattod.2008.08.039


July 2014 192

Huber, G. W., & Corma, A. (2007). Synergies between bio- and oil refineries for the production of fuels from biomass. Angewandte Chemie (International Ed. in English), 46(38), 7184–201. doi:10.1002/anie.200604504

Huber, G. W., Cortright, R. D., & Dumesic, J. a. (2004). Renewable alkanes by aqueous-phase reforming of biomass-derived oxygenates. Angewandte Chemie (International Ed. in English), 43(12), 1549–51. doi:10.1002/anie.200353050

Huber, G. W., & Dumesic, J. a. (2006). An overview of aqueous-phase catalytic processes for production of hydrogen and alkanes in a biorefinery. Catalysis Today, 111(1-2), 119–132. doi:10.1016/j.cattod.2005.10.010

Huber, G. W., Iborra, S., & Corma, A. (2006). Synthesis of transportation fuels from biomass: chemistry, catalysts, and engineering. Chemical Reviews, 106(9), 4044–4098. doi:10.1021/cr068360d

Humbird, D., Davis, R., Tao, L., Kinchin, C., Hsu, D., Aden, A., … Dudgeon, D. (2011). Process design and economics for biochemical conversion of lignocellulosic biomass to ethanol (p. 147). Golden, CO, USA: National Renewable Energy Laboratory.

IATA. (2013). Fact Sheet: Alternative Fuels (p. 2). Retrieved from https://www.iata.org/pressroom/facts_figures/fact_sheets/pages/alt-fuels.aspx

IEA. (2011a). Medium Term Oil &Gas Markets 2011 (p. 272). Paris: OECD Publishing.

IEA. (2011b). Technology Roadmap: Biofuels for Transport (p. 56). Paris: OECD Publishing. doi:10.1787/9789264118461-en

IEA. (2012a). Medium-Term Oil Market Report 2012 (p. 144). Paris: OECD Publishing. doi:10.1787/9789264188471-en

IEA. (2012b). World Energy Outlook 2012. International Energy Agency (IEA), Paris (p. 690). Paris. Retrieved from http://203.117.10.102/media/news_pdfs/WEO2012 Singapore Fatih Birol.pdf

IEA Bioenergy. (2007). Black liquor gasification. In Summary and Conclusions from the IEA Bioenergy ExCo54 Workshop (Vol. 64, p. 12). doi:10.1002/cjce.5450640221

IMO. (2012). North American emission control area comes into effect on 1 August 2012. International Maritime Organization Press Briefings. Retrieved June 04, 2013, from http://www.imo.org/mediacentre/pressbriefings/pages/28-eca.aspx

IndexMundi. (2013). Commodity Prices - Price Charts, Data, and News - IndexMundi. Retrieved August 07, 2013, from http://www.indexmundi.com/commodities/

INEOS. (2013). INEOS. official website. Retrieved August 15, 2013, from http://energy.gov/articles/turning-waste-fuel-how-ineos-biorefinery-changing-clean-energy-game


July 2014 193

Informa Economics. (2013). BIO-BUTANOL: THE GAME CHANGER, An Emerging Biofuel and Biochemical. A Multi-Client Study in Partnership with Chemical Strategies Group (p. 17). Retrieved from http://www.informaecon.com/MCSBiobutanol2013.pdf

Ingledew, W. M. (2008). Starch-Based Ethanol Production. Retrieved August 28, 2013, from http://www1.agric.gov.ab.ca/$department/deptdocs.nsf/all/crop12128

InvestmentMine. (2013). Commodity and Metal Prices, Metal Price Charts - InvestmentMine. Retrieved August 07, 2013, from http://www.infomine.com/investment/metal-prices/

Jang, Y.-S., Malaviya, A., Cho, C., Lee, J., & Lee, S. Y. (2012). Butanol production from renewable biomass by clostridia. Bioresource Technology, 123, 653–63. doi:10.1016/j.biortech.2012.07.104

Jin, C., Yao, M., Liu, H., Lee, C. F., & Ji, J. (2011). Progress in the production and application of n-butanol as a biofuel. Renewable and Sustainable Energy Reviews, 15(8), 4080–4106. doi:10.1016/j.rser.2011.06.001

Jones, S. B., Valkenburg, C., Walton, C. W., Elliott, D. C., Holladay, J. E., Stevens, D. J., … Czernik, S. (2009). Production of Gasoline and Diesel from Biomass via Fast Pyrolysis , Hydrotreating and Hydrocracking : A Design Case, (February).

Jovanovic Tews, I., Jones, S. B., Santosa, D. M., Dai, Z., Ramasamy, K., & Zhu, Y. (2010). A Survey of Opportunities for Microbial Conversion of Biomass to Hydrocarbon Compatible Fuels .

Khosla, V. (2011). What matters in Biofuels & where are we? Green Tech Media. Retrieved from https://www.greentechmedia.com/articles/read/guest-post-vinod-khosla-on-what-matters-in-biofuels

King, A. (2012). Ultra-low sulfur jet fuel on the radar. RSC Chemistry World. Retrieved June 04, 2013, from http://www.rsc.org/chemistryworld/2012/05/ultra-low-sulfur-jet-fuel-radar

KiOR. (2013). KiOR, Inc. - official website. various sections. Retrieved August 07, 2013, from http://www.kior.com/

KIT. (2013, February 28). bioliq® Pilot Plant: Successful Operation of High-pressure Entrained Flow Gasification. Press releases. Jutta Sendelbach. Retrieved August 28, 2013, from http://www.kit.edu/visit/pi_2013_12692.php

Knothe, G. (2011). A technical evaluation of biodiesel from vegetable oils vs. algae. Will algae-derived biodiesel perform? Green Chemistry, 13(11), 3048. doi:10.1039/c0gc00946f

Kolstad, J. J., Vink, E. T. H., De Wilde, B., & Debeer, L. (2012). Assessment of anaerobic degradation of IngeoTM polylactides under accelerated landfill conditions. Polymer Degradation and Stability, 97(7), 1131–1141. doi:10.1016/j.polymdegradstab.2012.04.003

Landalv, I. (2013). Status report on Demonstration Plants for Advances Biofuels Production - Thermochemical Pathways. Retrieved from http://www.biofuelstp.eu/spm5/pres/landalv.pdf


July 2014 194

Lane, J. (2011a). Highwater Ethanol is first in Butamax Early Adopters Group, seeks biobutanol options. Biofuels Digest. Retrieved August 28, 2013, from http://www.biofuelsdigest.com/bdigest/2011/12/05/highwater-ethanol-is-first-in-butamax-early-adopters-group-seeks-biobutanol-options/

Lane, J. (2011b). The Range Fuels failure : Biofuels Digest. Biofuels Digest. Retrieved August 28, 2013, from http://www.biofuelsdigest.com/bdigest/2011/12/05/the-range-fuels-failure/

Lane, J. (2012). Advanced Biofuels and chemicals The race for scale 2012-2025 1300 + refineries by 2025 (p. 6). Retrieved from https://www.google.ca/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&ved=0CC4QFjAA&url=http%3A%2F%2Fwww.ascension-publishing.com%2FBIZ%2FWBM-0312-1.pdf&ei=i3nRUpPxDMTboASztIHwBg&usg=AFQjCNH37e25uRH6h4EnfP14SAVE5JW4mg&sig2=PrvCkoDPGYnXcH3Dl9FjYw&bvm=bv.59026428,d.cGU

Lane, J. (2013). KiOR commences shipments of cellulosic gasoline : Biofuels Digest. Retrieved August 07, 2013, from http://www.biofuelsdigest.com/bdigest/2013/07/02/kior-commences-shipments-of-cellulosic-gasoline/

Lanzatech. (2013). Lanzatech - official website. various sections. Retrieved August 24, 2013, from www.lanzatech.com

Lau, F., Bowen, D., Zabransky, R. R., Slimane, R., & Doong, S. (2003). Techno-economic analysis of hydrogen production by gasification of biomass. Department of Energy … (p. 4). Des Plaines, IL, USA. Retrieved from http://www.eere.energy.gov/hydrogenandfuelcells/pdfs/32405b5.pdf

Lappas, a. a., Bezergianni, S., & Vasalos, I. a. (2009). Production of biofuels via co-processing in conventional refining processes. Catalysis Today, 145(1-2), 55–62. doi:10.1016/j.cattod.2008.07.001

Lee, S. Y., Park, J. H., Jang, S. H., Nielsen, L. K., Kim, J., & Jung, K. S. (2008). Fermentative butanol production by Clostridia. Biotechnology and Bioengineering, 101(2), 209–28. doi:10.1002/bit.22003

Lehto, J., Oasmaa, A., Solantausta, Y., Kytö, M., & Chiaramonti, D. (2013). Fuel oil quality and combustion of fast pyrolysis bio-oils. vtt.fi. Retrieved from http://www.vtt.fi/inf/pdf/technology/2013/T87.pdf

Levin, D. B., & Chahine, R. (2010). Challenges for renewable hydrogen production from biomass. International Journal of Hydrogen Energy, 35(10), 4962–4969. doi:10.1016/j.ijhydene.2009.08.067

Lin, Y., Li, C., Wan, H., Lee, H., & Liu, C. (2011). Catalytic Hydrodeoxygenation of Guaiacol on Rh-Based and Sulfided CoMo and NiMo Catalysts. Energy & Fuels, 890–896. doi:dx.doi.org/10.1021/ef101521z

Linde Engineering. (2012). http://www.the-linde-group.com/en/news_and_media/press_releases/news_120209.html. Press releases. Retrieved July 16, 2013, from http://www.the-linde-group.com/en/news_and_media/press_releases/news_120209.html


July 2014 195

Liu, J., Fan, L. T., Seib, P., Friedler, F., & Bertok, B. (2004). Downstream process synthesis for biochemical production of butanol, ethanol, and acetone from grains: generation of optimal and near-optimal flowsheets with conventional operating units. Biotechnology Progress, 20(5), 1518–27. doi:10.1021/bp049845v

LS9. (2013). LS9 official website. various sections. Retrieved August 25, 2013, from www.ls9.com

Lux Research. (2010). Bio-based Fuels and Materials through 2015: Growing Capacity Past a Drop in the Bucket. Retrieved from https://portal.luxresearchinc.com/research/report_excerpt/7015#executivesummary

Mabee, W. (2007). Policy options to support biofuel production. Biofuels, (April), 329–357. doi:10.1007/10

Machado, C. (2010). Technical characteristics and current status of butanol production and use as biofuel (p. 23). Retrieved from http://www.olade.org/biocombustibles/Documents/Ponencias Chile/Sesion 5_C Machado_Enbrapa_Brasil.pdf

Maniatis, K. (2013). Personal email comminucation on 23 Decenber 2013, with K.Maniatis, European Commission, DG Energy.

Maniatis, K., Weitz, M., & Zschoke, A. (2011). 2 million tons per year : A performing biofuels supply chain for EU aviation (pp. 1–27).

Marchetti, J.M., Miguel, V.U., & Errazu A.F. (2008) Techno-economic study of different alternatives for biodiesel production, Fuel Processing Technology, 89:740–748

Marker, T. (2005). Opportunities for biorenewables in oil refineries. Final Technical Report.

Marker, T. L., Felix, L. G., Linck, M. B., & Roberts, M. J. (2012). Integrated Hydropyrolysis and Hydroconversion ( IH 2 ) for the Direct Production of Gasoline and Diesel Fuels or Blending Components from Biomass , Part 1 : Proof of Principle Testing, 31(2), 191–199. doi:10.1002/ep

Meehan, P. M. (2009). Investigations into the fate and behavior of selected inorganic compounds during biomass gasification by. Iowa State Univeristy. Retrieved from http://lib.dr.iastate.edu/cgi/viewcontent.cgi?article=1727&context=etd

Meier, D., Beld, B. Van De, Bridgwater, A. V, Elliott, D. C., Oasmaa, A., & Preto, F. (2013). State-of-the-art of fast pyrolysis in IEA bioenergy member countries. Renewable and Sustainable Energy Reviews, 20, 619–641. doi:10.1016/j.rser.2012.11.061

Milne, T., Abatzoglou, N., & Evans, R. (1998). Biomass gasifier“ tars”: their nature, formation, and conversion (p. 204). Golden, CO, USA. Retrieved from http://www.ps-survival.com/PS/Gasifiers/Biomass_Gasifier_Tars_Their_Nature_Formation_And_Conversion_1998.pdf


July 2014 196

Mohamed, M. a., Soelem, R., Attar, F., & Ozalp, N. (2011). Hydrogen Production and Utilization in Petroleum Refineries: A Study of the U.S. Oil and Gas Industry. In ASME 2011 5th International Conference on Energy Sustainability, Parts A, B, and C (pp. 903–916). Asme. doi:10.1115/ES2011-54752

Mortensen, P. M., Grunwaldt, J.-D., Jensen, P. a., Knudsen, K. G., & Jensen, a. D. (2011). A review of catalytic upgrading of bio-oil to engine fuels. Applied Catalysis A: General, 407(1-2), 1–19. doi:10.1016/j.apcata.2011.08.046

Muellerlanger, F., Tzimas, E., Kaltschmitt, M., & Peteves, S. (2007). Techno-economic assessment of hydrogen production processes for the hydrogen economy for the short and medium term. International Journal of Hydrogen Energy, 32(16), 3797–3810. doi:10.1016/j.ijhydene.2007.05.027

Munasinghe, P. C., & Khanal, S. K. (2010). Biomass-derived syngas fermentation into biofuels: Opportunities and challenges. Bioresource Technology, 101(13), 5013–22. doi:10.1016/j.biortech.2009.12.098

NABC. (2013). National Advanced Biofuels Consortium, USA. official website. Retrieved July 27, 2013, from http://www.nabcprojects.org/about.html

Narula, C. K., Davison, B. H., & Keller, M. (2012). Catalytic conversion of bio-alcohols to hydrocarbons Inventors : Technology - Platform for converting bioalcohols to hydrocarbon blend-stocks. ORNL. Retrieved from http://web.ornl.gov/adm/partnerships/events/SPARK_Dec12/Presentations/04_Catalytic Conversion-Bio-alcohols to Hydrocarbons-Speck.pdf

Neste Oil. (2013a). Neste Oil Production Capacity. Renewable Fuels. Retrieved January 06, 2014, from http://www.nesteoil.com/default.asp?path=1,41,11991,22708,22720

Neste Oil. (2013b). Neste Oil’s Interim Report for January-March 2013. Retrieved May 21, 2013, from http://www.nesteoil.com/default.asp?path=1,41,540,1259,1261,20618,21191

NETL. (2013). Gasification Plant Databases. Retrieved August 28, 2013, from http://www.netl.doe.gov/technologies/coalpower/gasification/worlddatabase/

Nexant. (2012). Bio-derived Jet Fuel Cost Benchmarking, Nexant report prepared for Gevo (p. 35).

Nielsen, R. P., Olofsson, G., & Søgaard, E. G. (2012). CatLiq – High pressure and temperature catalytic conversion of biomass: The CatLiq technology in relation to other thermochemical conversion technologies. Biomass and Bioenergy, 39, 399–402. doi:10.1016/j.biombioe.2012.01.035

NRC Canada. (2012). NRC flies the world’s first civil jet powered by 100 percent biofuel. News Releases. Retrieved August 22, 2013, from http://www.nrc-cnrc.gc.ca/eng/news/releases/2012/biofuels.html

NSF. (2011). Developing New Paradigms for Biofuel Separations to Enable an Alternative Fuels Future (Based on the April 4-5, 2011 Workshop at Arlington, VA) (p. 27). Arlington, VA, USA.


July 2014 197

Nummisalo, V. (2012, December). Standardisation of fast pyrolysis bio-oil under CEN. IEA Bioenergy Task 34 Newsletter #32, (December), 15–16. Retrieved from http://www.pyne.co.uk/Resources/user/PyNe Issue 32 - December 2012 - website.pdf

Oasmaa, A., & Czernik, S. (1999). Fuel Oil Quality of Biomass Pyrolysis Oils: State of the Art for the End Users. Energy & Fuels, 13, 914–921.

Pearlson, M. N. (2011). A techno-economic and environmental assessment of hydroprocessed renewable distillate fuels (Master Thesis). Massachusetts Institute of Technology. Retrieved from http://dspace.mit.edu/handle/1721.1/65508

Peralta-Yahya, P. P., & Keasling, J. D. (2010). Advanced biofuel production in microbes. Biotechnology Journal, 5(2), 147–62. doi:10.1002/biot.200900220

Pfromm, P. H., Amanor-Boadu, V., Nelson, R., Vadlani, P., & Madl, R. (2010). Bio-butanol vs. bio-ethanol: A technical and economic assessment for corn and switchgrass fermented by yeast or Clostridium acetobutylicum. Biomass & Bioenergy, 34(4), 515–524. Retrieved from http://www.sciencedirect.com/science/article/pii/S0961953409002645

Phillips, S., Tarud, J., Biddy, M., & Dutta, A. (2011). Gasoline from wood via integrated gasification, synthesis, and methanol-to-gasoline technologies, (January). Retrieved from http://www.osti.gov/greenenergy/rddetail?osti_id=1004790

Plattts. (2013). Shell mulls more efficient, but still costly GTL mega-project on US Gulf - Oil | Platts News Article & Story. Retrieved August 28, 2013, from http://www.platts.com/latest-news/oil/london/shell-mulls-more-efficient-but-still-costly-gtl-26061813

Pray, T. (2010). Amyris presentation at the DOE Biomass R&D technical advisroy comittee: Drop-in Fuels Panel (p. 20). Aurora, Colorado, USA. Retrieved from http://www.biomassboard.gov/pdfs/biomass_tac_todd_pray_09_29_2010.pdf

Qureshi, N., & Blaschek, H. P. (1999). Production of acetone butanol ethanol (ABE) by a hyper-producing mutant strain of Clostridium beijerinckii BA101 and recovery by pervaporation. Biotechnology Progress, 15(4), 594–602. doi:10.1021/bp990080e

Ranjan, A., & Moholkar, V. S. (2012). Biobutanol : science , engineering , and economics, (December 2011), 277–323. doi:10.1002/er

Repotec. (2013). Repotec-official website. Technology. Retrieved July 15, 2013, from http://www.repotec.at/

Ringer, M., Putsche, V., & Scahill, J. (2006). Large-Scale Pyrolysis Oil Production : A Technology Assessment and Economic Analysis (p. 93). Golden, CO, USA. Retrieved from http://www.nrel.gov/docs/fy07osti/37779.pdf

RISI. (2013). Luleå University of Technology takes over Chemrec and its plant for gasification of green fuels in Piteå, Sweden. TECH CHANNEL. Retrieved August 28, 2013, from


July 2014 198

http://www.risiinfo.com/techchannels/powerenergy/Lulea-University-of-Technology-takes-over-Chemrec-and-its-plant-for-gasification-of-green-fuels-in-Pitea.html

Ro, D.-K., Paradise, E. M., Ouellet, M., Fisher, K. J., Newman, K. L., Ndungu, J. M., … Keasling, J. D. (2006). Production of the antimalarial drug precursor artemisinic acid in engineered yeast. Nature, 440(7086), 940–3. doi:10.1038/nature04640

Roig, S. (2012, July). U.S. Navy’s “Great Green Fleet” debuts in Pacific. Reuters News. Retrieved from http://uk.reuters.com/article/2012/07/19/uk-usa-navy-greenfleet-idUKBRE86I0C320120719

Rosillo-Calle, F., Thrän, D., Seiffert, M., & Teelucksingh, S. (2012). The Potential Role Of Biofuels In Commercial Air Transport–Biojetfuel. IEA bioenergy Task 40. Retrieved from http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:The+Potential+and+Role+of+Biofuels+in+Commercial+Air+Transport+-+Biojetfuel#0

RSB. (2012). Indirect Impacts of biofuel production and the RSB Standard. Retrieved from http://bioenergywiki.net/images/e/ec/12-04-13_RSB_Indirect_Impacts.pdf

Rude, M. A., & Schirmer, A. (2009). New microbial fuels: a biotech perspective. Current Opinion in Microbiology, 12(3), 274–281. Retrieved from http://dx.doi.org/10.1016/j.mib.2009.04.004

S&T2. (2013). Crude oil production updates (p. 80). Delta, BC, Canada. Retrieved from http://www.ghgenius.ca/reports.php

SAFN. (2011). Sustainable Aviation Fuels Northwest, Powering the next generation of flight (pp. 1–131). Retrieved from http://www.safnw.com/sustainable-aviations-fuels-bibliography/

SAFUG. (2014). Sustainable Aviation Fuel Users Group - official website. Case studies. Retrieved January 06, 2014, from http://www.safug.org/case-studies/

Sasol. (2013). Overview | Sasol. Retrieved August 28, 2013, from http://www.sasol.com/innovation/gas-liquids/overview

Sawaya Jank, M. (2011). Biofuel Industry: A local and global view. Presentation at the BBEST conference in Campos de Jordao, Brazil on August 15th 2011. Retrieved from http://www.bbest.org.br/2011/images/palestras/marcos_jank

Schnepf, R., & Yacobucci, B. D. (2013). Renewable Fuel Standard ( RFS ): Overview and Issues (p. 35). Washington, DC, USA. Retrieved from http://www.fas.org/sgp/crs/misc/R40155.pdf

Sims, R., Taylor, M., Saddler, J., & Mabee, W. (2008). From 1st - to 2nd - generation biofuel technologies, An overview of current industry and RD&D activities (p. 124). Paris. Retrieved from http://www.task39.org/LinkClick.aspx?fileticket=+KPhJZy8YAI=&tabid=4426&language=en-US

Sitzmann, J. (2009, October). Upgrading of fast pyrolysis oils by hot filtration. Aston University, UK. Retrieved from http://eprints.aston.ac.uk/15742/


July 2014 199

Solantausta, Y. (2011). BIOCOUP, Co-processing of upgraded bio-liquids in standard refinery units, Final Activity Report (pp. 1–38). Espoo, Finland. Retrieved from http://www.biocoup.com/uploads/media/BIOCOUP_Publishable_final_activity_report_01.pdf

Solazyme. (2013). Solazyme - official website. various sections. Retrieved August 23, 2013, from www.solazyme.com

Sorda, G., Banse, M., & Kemfert, C. (2010). An overview of biofuel policies across the world. Energy Policy, 38(11), 6977–6988. doi:10.1016/j.enpol.2010.06.066

Sotelo-Boyas, R., Trejo-Zarraga, F., & Hernandez-Loyo, F. de J. (2012). Hydroconversion of Triglycerides into Green Liquid Fuels, Hydrogenation,. In I. Karamé (Ed.), Hydrogenation. Mexico City. doi:10.5772/48710

Speight, G. S., & Ancheyta, J. (2007). Hydroprocessing of heavy oils and residua. (G. S. Speight & J. Ancheyta, Eds.) (p. 345). CRC / Taylor & Francis.

Speight, J. (2006). The Chemistry and Technology of Petroleum, Fourth Edition (4th ed., p. 984). Boca Raton, FL, USA: CRC Press. Retrieved from http://books.google.ca/books?id=ymL2S9RWzx4C&pg=PA190&lpg=PA190&dq=oxygen+content+in+petroleum&source=bl&ots=buwkrrFYLi&sig=TZEK3JoNSmQA3HIKbAnN4QqSQUE&hl=en&sa=X&ei=sG72UqrqMeWFyQGuk4GIBA&ved=0CCkQ6AEwAA#v=onepage&q=oxygen content in petroleum&f=false

Speight, J., & Özüm, B. (2002). Petroleum refining processes (p. 706). New York, NY, USA: Marcel Dekker Inc. Retrieved from http://books.google.com/books?hl=en&lr=&id=4ciEKGUOyksC&oi=fnd&pg=PP1&dq=Petroleum+Refining+Processes&ots=R4w4ok15dU&sig=M0fznbD-gzmbQcj5gp3_E1ZOaE4

Starck, J. (2012, July). Green Fuel Nordic. IEA Bioenergy Task 34 Newsletter #32, 5–6. Retrieved from http://www.pyne.co.uk/Resources/user/PyNe July 2012 - Issue 31.pdf

Stephen, J. D., Mabee, W. E., & Saddler, J. (2010). Biomass logistics as a determinant of second- generation biofuel facility scale , location and technology. Biofuels Bioproducts & Bioerefining, 4(5), 513–518. doi:10.1002/bbb

Stratiev, D., Tzingov, T., Shishkova, I., & Dermatova, P. (2009). Hydrotreating units chemical hydrogen consumption analysis: A tool for improving refinrey hydrogen management. In 44th International Petroleum Conference, Bratislava, Slovak Republic, September 21-22, 2009 HYDROTREATING (pp. 1–14). Retrieved from http://www.vurup.sk/sites/vurup.sk/archivedsite/www.vurup.sk/conferences/44_ipc/44ipc/Sustainable Crude Oil Processing_Tudays and Future Technologies, the role of the catalysis/posters/Hydrotreating Units Chemical Hydrogen Consumption Analysis_Stratiev.pdf

Swanson, R. M., Platon, A., Satrio, J. a., & Brown, R. C. (2010). Techno-economic analysis of biomass-to-liquids production based on gasification. Fuel, 89, S11–S19. doi:10.1016/j.fuel.2010.07.027


July 2014 200

Swanson, R., Satrio, J., Brown, R., Platon, A., & Hsu, D. (2010). Techno-economic analysis of biofuels production based on gasification. Contract (p. 165). Golden, CO, USA. Retrieved from http://www.nrel.gov/docs/fy11osti/46587.pdf

Trench, C. J. (2001). How Pipelines Make the Oil Market Work – Their Networks, Operation and Regulation (p. 22). New York, NY, USA. Retrieved from http://www.iatp.org/documents/how-pipelines-make-the-oil-market-work-their-networks-operation-and-regulation

Tyner, W. E. (2010). Why the push for drop-in biofuels? Biofuels, 1(6), 813–814.

Tyner, W. E., & Viteri, D. (2010). Policy Update: Implications of blending limits on the US ethanol and biofuels markets. Biofuels, 1(2), 251–253. doi:10.4155/bfs.09.24

USDA (2010). A USDA Regional Roadmap to Meeting the Biofuels Goals of the Renewable Fuels Standard by 2022 (p. 21). Retrieved June 15, 2013 from http://www.usda.gov/documents/USDA_Biofuels_Report_6232010.pdf

USDA (2012). Agriculture and Aviation: Partners in Prosperity (p. 49). Retrieved June 17, 2013 from http://www.usda.gov/documents/usda-farm-to-fly-report-jan-2012.pdf

USDA (2013). Major Vegetable Oils: World Supply and Distribution. Foreign Agricultural Service statistics database. Retrieved May 20, 2013 from http://apps.fas.usda.gov/psdonline/psdReport.aspx?hidReportRetrievalName=Table+03%3a+Major+Vegetable+Oils%3a+World+Supply+and+Distribution+(Commodity+View)++++++++++++++++++++++&hidReportRetrievalID=533&hidReportRetrievalTemplateID=5

U.S. EIA. (2006). Changing Trends in the Refining Industry. Independant statistics and analysis, U.S. Energy Information Administration. Retrieved June 10, 2013, from http://www.eia.gov/oiaf/aeo/otheranalysis/aeo_2006analysispapers/tri.html

U.S. EIA. (2007). Refinery outages, description and potential impact on petroleum product prices (Vol. SR/OOG/200, pp. 1–53). Washington, DC: Energy Information Administration. US Department of Energy. Retrieved from http://www.eia.gov/oiaf/servicerpt/refinery_outages/sroog200701.pdf

U.S. EIA. (2010). International Energy Statistics. Independant statistics and analysis, U.S. Energy Information Administration. Retrieved June 09, 2013, from http://www.eia.gov/cfapps/ipdbproject/iedindex3.cfm?tid=5&pid=62&aid=2&cid=regions&syid=2006&eyid=2010&unit=TBPD

US EPA. (1995). Profile of the petroleum refining industry (p. 146). Washington, DC, USA. Retrieved from http://books.google.com/books?hl=en&lr=&id=nlxEAAAAYAAJ&oi=fnd&pg=PP10&dq=Profile+of+the+Petroleum+Refining+Industry&ots=QdIbMEq1Oi&sig=xx1Rl6TzoYYcfdRCZam-hHNmgn8

US EPA. (2013, November 15). Renewable Fuel Standard (RFS). Fuels and Fuel Additives, U.S. Envioronmental Protection Agency. Retrieved June 09, 2013, from http://www.epa.gov/otaq/fuels/renewablefuels/index.htm


July 2014 201

Van der Drift, B. (2013). Biomass gasification in the Netherlands: IEA Bioenergy Task 33 Report (p. 33). Retrieved from http://www.ieatask33.org/app/webroot/files/file/country_reports/NL_July2013.pdf

Van Noorden, R. (2013, September). European Parliament votes to limit crop-based biofuels. Nature News Blog. Retrieved from European Parliament votes to limit crop-based biofuels

Vane, L. M., Alvarez, F. R., Rosenblum, L., & Govindaswamy, S. (2013). Efficient recovery and dehydration of 1-butanol, acetone/butanol/ethanol (ABE), and isobutanol from dilute aqueous solutions using hybrid vapor tripping-vapor permeation process. In 35th Symposium on Biotechnology for Fuels and Chemicals.

Värmlandsmetanol. (2012). World ’ s first Commercial Scale Biomethanol Plant in Hagfors SWEDEN Biomass as received District heat export Methanol energy (p. 11). Retrieved from http://www.varmlandsmetanol.se/dokument/History March 12.pdf

Venderbosch, R., & Prins, W. (2011). Fast Pyrolysis. In R. C. Brown (Ed.), Thermochemical Processing of Biomass: Conversion into Fuels, Chemicals and Power (1st ed., pp. 144–176). Hoboken, NJ, USA: Wiley-VCH. Retrieved from http://onlinelibrary.wiley.com/book/10.1002/9781119990840

Vennestrøm, P. N. R., Osmundsen, C. M., Christensen, C. H., & Taarning, E. (2011). Beyond petrochemicals: the renewable chemicals industry. Angewandte Chemie (International Ed. in English), 50(45), 10502–9. doi:10.1002/anie.201102117

Vermeire, M. B. (2012). Everything You Need to Know About Marine Fuels (p. 32). San Ramon, CA, USA. Retrieved from http://www.chevronmarineproducts.com/docs/Chevron_EverythingYouNeedToKnowAboutFuels_v3_1a_DESKTOP.pdf

Vispute, T. P., Zhang, H., Sanna, A., Xiao, R., & Huber, G. W. (2010). Renewable chemical commodity feedstocks from integrated catalytic processing of pyrolysis oils. Science (New York, N.Y.), 330(6008), 1222–7. doi:10.1126/science.1194218

Walker, G. M., & Brew, J. I. (2011). 125th Anniversary Review : Fuel Alcohol : Current Production and Future Challenges, 117(1), 3–22.

Wander, P. R., Altafini, C. R., & Barreto, R. M. (2004). Assessment of a small sawdust gasification unit. Biomass and Bioenergy, 27(5), 467–476. doi:10.1016/j.biombioe.2004.04.003

Wang, H., Male, J., & Wang, Y. (2013). Recent Advances in Hydrotreating of Pyrolysis Bio-Oil and Its Oxygen-Containing Model Compounds. ACS Catalysis, 3(5), 1047–1070. doi:10.1021/cs400069z

Weber, C., Farwick, A., Benisch, F., Brat, D., Dietz, H., Subtil, T., & Boles, E. (2010). Trends and challenges in the microbial production of lignocellulosic bioalcohol fuels. Applied Microbiology and Biotechnology, 87(4), 1303–15. doi:10.1007/s00253-010-2707-z


July 2014 202

Wender, I. (1996). Reactions of synthesis gas. Fuel Processing Technology, 48(I 996), 189–297. Retrieved from http://www.sciencedirect.com/science/article/pii/S037838209601048X

Weyen, R. (2012, March). A Refiner ’ s Perspective on Advanced Biofuels. NABC Highlights, 4. Retrieved from http://www.nabcprojects.org/pdfs/refiner_perspective_advanced_biofuels.pdf

Wildschut, B. J. (n.d.). Pyrolysis Oil Upgrading to Transportation Fuels by Catalytic Hydrotreatment.

Wildschut, J., Iqbal, M., Mahfud, F. H., Cabrera, I. M., Venderbosch, R. H., & Heeres, H. J. (2010). Insights in the hydrotreatment of fast pyrolysis oil using a ruthenium on carbon catalyst. Energy & Environmental Science, 3, 962–970. Retrieved from http://pubs.rsc.org/en/content/articlehtml/2010/EE/B923170F

Wildschut, J., Mahfud, F. H., Venderbosch, R. H., & Heeres, H. J. (2009). Hydrotreatment of Fast Pyrolysis Oil Using Heterogeneous Noble-Metal Catalysts. Industrial & Engineering Chemistry, 48(23), 10324–10334.

Wildschut, J., Melian-Cabrera, I., & Heeres, H. J. (2010). Catalyst studies on the hydrotreatment of fast pyrolysis oil. Applied Catalysis B: Environmental, 99(1-2), 298–306. Retrieved from http://www.sciencedirect.com/science/article/pii/S0926337310002869

Wright, M. M., Daugaard, D. E., Satrio, J. a., Brown, R. C., & Hsu, D. D. (2010). Techno-Economic Analysis of Biomass Fast Pyrolysis to Transportation Fuels. Fuel, 89(November), S2–S10. doi:10.1016/j.fuel.2010.07.029

Yerushalmi, L., Volesky, B., & Szczesny, T. (1985). Effect of increased hydrogen partial pressure on the acetone-butanol fermentation by Clostridium acetobutylicum. Applied Microbiology and Biotechnology, 22(2), 103–107. doi:10.1007/BF00250028

Zhang, H., Cheng, Y.-T., Vispute, T. P., Xiao, R., & Huber, G. W. (2011). Catalytic conversion of biomass-derived feedstocks into olefins and aromatics with ZSM-5: the hydrogen to carbon effective ratio. Energy & Environmental Science, 4(6), 2297. doi:10.1039/c1ee01230d
